Skip to content

Artificial Intelligence

LangChain
API integration

Ship Artificial Intelligence features without building the integration. Full LangChain API access via Proxy and 550+ MCP-ready tools for AI agents — extend models and mappings to fit your product.

Talk to us
LangChain

Use Cases

Why integrate with LangChain

Common scenarios for SaaS companies building LangChain integrations for their customers.

01

Embed LLM observability into your AI-powered SaaS

Let your customers connect their LangSmith workspace so traces, runs, and thread-level analytics from your AI features flow into their existing LLMOps stack. Removes the need for them to manually wire up SDKs or maintain duplicate observability infrastructure.

02

Close the feedback loop from your product to prompt engineers

Capture thumbs-up/down, edits, and corrections your end users make on AI-generated content and push them as structured feedback tied to LangSmith run IDs. Gives your customers' AI teams real production signal to fine-tune prompts and models.

03

Route low-confidence AI outputs into human review queues

When your AI agent fails or produces low-confidence responses, automatically push those runs into your customer's LangSmith annotation queues for human-in-the-loop grading. Lets QA and prompt teams triage failures without leaving their existing review workflow.

04

Surface AI cost and usage inside FinOps dashboards

Pull granular billing, token usage, and thread stats from LangSmith into your financial analytics or cloud cost platform. Enterprises can then unify LLM spend with the rest of their infrastructure budget and set programmatic usage limits.

05

Sync ground-truth datasets between your app and LangSmith

Let users promote real interactions from your platform (support tickets, generated docs, chat transcripts) into LangSmith datasets as test examples. Makes it trivial for your customers to build evaluation suites from actual production edge cases.

What You Can Build

Ship these features with Truto + LangChain

Concrete product features your team can ship faster by leveraging Truto’s LangChain integration instead of building from scratch.

01

Two-way feedback sync on AI outputs

Post user ratings, comments, and corrections to LangSmith feedback endpoints keyed by run ID, and read existing feedback back to display quality signals in your UI.

02

Automated annotation queue population

Create annotation queues and push runs into them programmatically when your app detects hallucinations, low confidence, or negative user reactions.

03

Dataset and example management UI

Give users an in-app interface to create datasets, bulk-upload examples, manage dataset splits, and trigger dataset runs without opening the LangSmith console.

04

Embedded trace and thread explorer

Fetch runs, thread traces, and thread stats to render step-by-step agent execution, latency, and token cost views natively inside your product.

05

Workspace and API key provisioning

Let users create workspaces, mint scoped API keys or service keys, and configure TTL and usage limits per tenant during onboarding to your integration.

06

Prompt and experiment CI hooks

Trigger experiment runs, comparative experiments, and platform evaluators from your CI/CD or release workflow, then pull results back to gate deploys.

SuperAI

LangChain AI agent tools

Comprehensive AI agent toolset with fine-grained control. Integrates with MCP clients like Cursor and Claude, or frameworks like LangChain.

list_all_lang_chain_info_healths

Get health information about the current LangSmith deployment. Returns the HealthInfoGetResponse object containing deployment health attributes.

get_single_lang_chain_session_dashboard_by_id

Get a prebuilt dashboard for a tracing project (session) in LangSmith. Returns: title, description, index, id, session_id, charts, sub_sections, layout, created_at, updated_at. Required: session_id.

get_single_lang_chain_session_by_id

Get a specific tracer session (project) in LangSmith by id. Returns: start_time, end_time, extra, name, description, default_dataset_id, reference_dataset_id, trace_tier, id, run_count, latency_p50, latency_p99, first_token_p50, first_token_p99, total_tokens, prompt_tokens, completion_tokens, total_cost, prompt_cost, completion_cost, tenant_id, last_run_start_time, last_run_start_time_live,…. Required: id.

update_a_lang_chain_session_by_id

Update a tracer session (project) in LangSmith by id. Returns: start_time, end_time, extra, name, description, default_dataset_id, reference_dataset_id, trace_tier, id, tenant_id, last_run_start_time_live. Required: id.

delete_a_lang_chain_session_by_id

Delete a tracer session (project) in LangSmith by id. Returns an empty 204 response on success. Required: id.

create_a_lang_chain_session

Create a new tracer session (project) in LangSmith. Returns: start_time, end_time, extra, name, description, default_dataset_id, reference_dataset_id, trace_tier, id, tenant_id, last_run_start_time_live. Requires a JSON request body with session configuration.

lang_chain_sessions_bulk_delete

Delete multiple LangSmith tracer sessions (projects) in bulk by their session IDs. Returns a successful response with no documented response body. Required: session_ids.

list_all_lang_chain_sessions

List LangSmith tracer sessions (projects) with extensive filtering and sorting options. Returns: start_time, end_time, extra, name, description, default_dataset_id, reference_dataset_id, trace_tier, id, run_count, latency_p50, latency_p99, first_token_p50, first_token_p99, total_tokens, prompt_tokens, completion_tokens, total_cost, prompt_cost, completion_cost, tenant_id, last_run_start_time,…

list_all_lang_chain_metadata

List the top K metadata values for each key in a LangSmith tracer session. Returns a dictionary mapping each metadata key name to a list of its top K string values. Required: session_id.

list_all_lang_chain_session_views

List all filter views for a LangSmith session, optionally filtered by type. Returns: filter_string, trace_filter_string, tree_filter_string, thread_filter_string, display_name, description, type, syntax_version, start_time, end_time, duration, id, session_id, created_at, updated_at. Required: session_id.

create_a_lang_chain_session_view

Create a new filter view for a LangSmith session. Returns the created filter view including id, session_id, display_name, description, and type. Required: session_id, display_name.

update_a_lang_chain_session_view_by_id

Update an existing filter view in a LangSmith session. Returns the updated filter view including id, session_id, display_name, description, and type. Required: id, session_id.

delete_a_lang_chain_session_view_by_id

Delete a specific filter view from a LangSmith session by id. Returns an empty response on success. Required: id, session_id.

get_single_lang_chain_session_view_by_id

Get a specific filter view from a LangSmith session by id. Returns: filter_string, trace_filter_string, tree_filter_string, thread_filter_string, display_name, description, type, syntax_version, start_time, end_time, duration, id, session_id, created_at, updated_at. Required: id, session_id.

lang_chain_view_renames_bulk_update

Rename a filter view's display_name and description in LangSmith. Only display_name and description can be modified. Returns the updated filter view including id, display_name, description, created_at, and updated_at. Required: session_id, view_id, display_name.

list_all_lang_chain_session_insights

List all run clustering insight jobs for a LangSmith session. Returns each job including its id; the full response shape is defined by the upstream GetRunClusteringJobsResponse schema. Beta endpoint. Required: session_id.

create_a_lang_chain_session_insight

Create a new run clustering insights job for a LangSmith session. Returns: id, name, status, error, attributes. Beta endpoint. Required: session_id.

update_a_lang_chain_session_insight_by_id

Update an existing run clustering insights job for a LangSmith session. Returns: name, status. Beta endpoint. Required: id, session_id, name.

delete_a_lang_chain_session_insight_by_id

Delete a run clustering insights job for a LangSmith session. Returns: id, message, attributes. Beta endpoint. Required: id, session_id.

get_single_lang_chain_session_insight_by_id

Get a specific run clustering insights job for a LangSmith session. Returns: id, name, status, start_time, end_time, created_at, metadata, shape, error, config_id, clusters, report, attributes. Beta endpoint. Required: id, session_id.

list_all_lang_chain_insights_configs

List all insights job configs for a LangSmith tracing session. Returns the configs collection with each item containing an id and config-type-specific fields. This is a beta endpoint. Required: session_id.

create_a_lang_chain_insights_config

Create an insights job config for a LangSmith tracing session by saving a clustering job configuration. Returns the saved config object with an id and config-type-specific fields. This is a beta endpoint. Required: session_id, name, config.

update_a_lang_chain_insights_config_by_id

Update an existing insights job config for a LangSmith tracing session. Returns the updated config object with an id and config-type-specific fields. This is a beta endpoint. Required: id, session_id.

delete_a_lang_chain_insights_config_by_id

Delete an insights job config from a LangSmith tracing session. Returns an empty 204 response on success. This is a beta endpoint. Required: id, session_id.

create_a_lang_chain_configs_generate

Auto-generate an insights job config for a LangSmith session (Beta). Returns the generated insights job config object. Required: session_id, user_context.

get_single_lang_chain_insight_cluster_by_id

Get a specific run cluster from an insights job in LangSmith. Returns the cluster object with its id and schema-specific attributes defined in the GetRunClusterResponse schema. Required: id, session_id, job_id.

list_all_lang_chain_insight_runs

List runs from a LangSmith insights clustering job, optionally filtered by cluster. Returns: runs, offset. Beta endpoint. Required: session_id, job_id.

list_all_lang_chain_workspaces

List LangSmith workspaces visible to the current auth in the current org. Returns: id, organization_id, created_at, display_name, is_personal, is_deleted, tenant_handle, data_plane_url, read_only, role_id, role_name, permissions. Does not create a new workspace or organization.

create_a_lang_chain_workspace

Create Workspace. Returns: id, organization_id, created_at, display_name, is_personal, is_deleted, tenant_handle, data_plane_url. Required: display_name.

update_a_lang_chain_workspace_by_id

Update a workspace in LangSmith by id. Returns: id, organization_id, created_at, display_name, is_personal, is_deleted, tenant_handle, data_plane_url. Required: id, display_name.

delete_a_lang_chain_workspace_by_id

Delete a workspace by id in LangSmith. Returns an empty 204 response on success. Required: id.

get_single_lang_chain_workspace_by_id

Get a single workspace by id in LangSmith, scoped to the current org and identity. Returns: id, organization_id, created_at, display_name, is_personal, is_deleted, tenant_handle, data_plane_url, read_only, role_id, role_name, permissions. Required: id.

list_all_lang_chain_workspaces_stats

Get statistics for the current LangSmith workspace, optionally filtered by one or more tag value IDs. Returns a TenantStats object with workspace-level resource and usage statistics. Optional: tag_value_id.

list_all_lang_chain_workspaces_usage_limits

Get the current LangSmith workspace's usage limits info, including configured trace and extended retention limits. Returns: in_reject_set, usage_limit_type, tenant_limit.

list_all_lang_chain_workspaces_shareds

List all shared entities and their tokens for the current LangSmith workspace. Returns: entities.

lang_chain_workspaces_shareds_bulk_delete

Bulk unshare entities by share tokens for the current LangSmith workspace. Returns a 200 response on success.

list_all_lang_chain_workspaces_secrets

List the names of secrets configured for the current LangSmith workspace. Returns: key, webhooks. Secret values are never returned by any endpoint.

create_a_lang_chain_workspaces_secret

Upsert workspace secrets in LangSmith by submitting an array of objects each with a key and value. A null value deletes that key. Returns an empty response on success; secret values are never returned by any endpoint.

list_all_lang_chain_secrets_encrypteds

Get encrypted workspace secrets for use with Fleet and external services in LangSmith. Returns the InternalSecretsResponse object as data, whose fields are defined by the upstream LangSmith schema and not individually documented in the API reference. Required: service.

list_all_lang_chain_workspaces_tag_keys

List tag keys in the current LangSmith workspace. Returns: key, description, id, created_at, updated_at.

create_a_lang_chain_workspaces_tag_key

Create a new tag key in the current LangSmith workspace. Returns: key, description, id, created_at, updated_at. Required: key.

update_a_lang_chain_workspaces_tag_key_by_id

Update an existing tag key by id in the current LangSmith workspace. Returns: key, description, id, created_at, updated_at. Required: id.

delete_a_lang_chain_workspaces_tag_key_by_id

Delete a tag key by id in the current LangSmith workspace. Returns an empty response on success. Required: id.

get_single_lang_chain_workspaces_tag_key_by_id

Get a single tag key by id in the current LangSmith workspace. Returns: key, description, id, created_at, updated_at. Required: id.

list_all_lang_chain_tag_key_tag_values

List all tag values for a specific tag key in the current LangSmith workspace. Returns: value, description, id, tag_key_id, created_at, updated_at. Required: tag_key_id.

create_a_lang_chain_tag_key_tag_value

Create a new tag value under a specific tag key in the current LangSmith workspace. Returns: value, description, id, tag_key_id, created_at, updated_at. Required: tag_key_id, value.

update_a_lang_chain_tag_key_tag_value_by_id

Update a tag value by id under a specific tag key in the current LangSmith workspace. Returns: value, description, id, tag_key_id, created_at, updated_at. Required: id, tag_key_id.

delete_a_lang_chain_tag_key_tag_value_by_id

Delete a tag value by id under a specific tag key in the current LangSmith workspace. Returns an empty 200 response on success. Required: id, tag_key_id.

get_single_lang_chain_tag_key_tag_value_by_id

Get a single tag value by id under a specific tag key in the current LangSmith workspace. Returns: value, description, id, tag_key_id, created_at, updated_at. Required: id, tag_key_id.

list_all_lang_chain_workspaces_taggings

List taggings in the current LangSmith workspace, optionally filtered by tag value. Returns: tag_key, tag_key_id, tag_value, tag_value_id, resources, id, resource_type, resource_id, created_at, updated_at.

create_a_lang_chain_workspaces_tagging

Create a tagging in the current LangSmith workspace by assigning a tag value to a resource. Returns: tag_value_id, resource_type, resource_id, id, created_at, updated_at. Required: tag_value_id, resource_type, resource_id.

delete_a_lang_chain_workspaces_tagging_by_id

Delete a tagging in the current LangSmith workspace by id. Returns an empty 204 response on success. Required: id.

list_all_lang_chain_workspaces_tags

List all resource tags in the current LangSmith workspace, optionally filtered by resource type. Returns an array of tag key objects, each including its id, key, and associated tag values. The optional resource_type query parameter narrows results to tags applicable to a specific resource type.

list_all_lang_chain_tags_resources

List tags for multiple LangSmith resources in a single batch request. Returns a map keyed by resource UUID where each value is an array of TagKeyWithValuesAndTaggings objects containing tag key, tag values, and tagging associations.

list_all_lang_chain_audit_logs

Retrieve audit log records for a LangSmith organization in OCSF API Activity (Class UID: 6003) format, compatible with security monitoring and SIEM tools. Returns: cursor, items. Max 100 per page. Required: start_time, end_time.

list_all_lang_chain_audit_logs_operations

List all possible audit log operation names in LangSmith for populating audit log filter UI. Returns a response object with an operations array of operation name strings, independent of any operations actually recorded so far for the organization. These names can be used as filter values when querying audit logs via the audit-logs endpoint.

list_all_lang_chain_ttl_settings

List the configured TTL settings for a given tenant in LangSmith. Returns the TTLSettings object including id and attributes with content-type-specific fields. The exact field set is defined by the upstream TTLSettings schema; consult the LangSmith REST API reference for details.

lang_chain_ttl_settings_bulk_update

Upsert TTL settings for a given tenant in LangSmith. Returns the updated TTLSettings object including id and attributes with content-type-specific fields. A JSON request body conforming to the upstream UpsertTTLSettingsRequest schema is required. Required: default_trace_tier.

list_all_lang_chain_orgs_ttl_settings

List the configured TTL settings for a LangSmith organization, covering org-level and tenant-level trace retention configuration. Returns: tenant_id, default_trace_tier, apply_to_all_projects, id, organization_id, created_at, updated_at, configured_by, longlived_ttl_days, thread_id, metadata, status, config. No required parameters.

lang_chain_orgs_ttl_settings_bulk_update

Upsert TTL settings for a LangSmith organization, configuring trace data retention at the org and tenant level. Returns the upserted TTL settings object including id, created_at, and updated_at. Requires a JSON request body. Required: default_trace_tier.

list_all_lang_chain_examples_counts

Count examples in LangSmith matching the provided query filters. Returns an integer representing the total number of matching examples. Supports filtering by dataset, splits, metadata, full-text content, and custom filter expressions.

lang_chain_examples_counts_list_public

Count examples accessible via a shared dataset share token in LangSmith. Returns an integer representing the total number of matching examples. Supports filtering by metadata and custom filter expressions. Required: share_token.

get_single_lang_chain_example_by_id

Get a specific example in LangSmith by id. Returns: outputs, dataset_id, source_run_id, source_session_id, source_run_start_time, source_trace_id, metadata, inputs, created_at, id, name, modified_at, source_thread_id, attachment_urls. Required: id.

update_a_lang_chain_example_by_id

Update a specific example in LangSmith by id. Returns the updated example object including id, created_at, modified_at, name, dataset_id, inputs, outputs, and metadata. Required: id.

delete_a_lang_chain_example_by_id

Soft delete an example in LangSmith by id. Only deletes the example in the 'latest' version of the dataset. Returns an empty 204 response on success. Required: id.

create_a_lang_chain_example

Create a new example in LangSmith. Returns the created example object including id, created_at, modified_at, name, dataset_id, inputs, outputs, and metadata. Required: dataset_id.

lang_chain_examples_bulk_delete

Soft delete multiple examples in LangSmith by their IDs. Only deletes the examples in the 'latest' version of the dataset. Returns an empty 204 response on success. Required: example_ids.

lang_chain_examples_bulk_create

Create multiple dataset examples in bulk in LangSmith. Returns an array of created example objects including id, dataset_id, inputs, outputs, created_at, modified_at, and name.

lang_chain_examples_bulk_update

Update multiple LangSmith examples in bulk via a legacy endpoint. Returns the updated example objects including id, inputs, outputs, metadata, and dataset_id. For updates involving attachments, use the platform datasets examples endpoint instead.

list_all_lang_chain_examples

List LangSmith dataset examples with optional filtering by dataset, splits, full-text search, and metadata. Returns: outputs, dataset_id, source_run_id, source_session_id, source_run_start_time, source_trace_id, metadata, inputs, created_at, id, name, modified_at, source_thread_id, attachment_urls, error. Max 100 per page.

lang_chain_examples_list_public

List LangSmith dataset examples shared via a public share token. Returns: outputs, dataset_id, source_run_id, source_session_id, source_run_start_time, source_trace_id, metadata, inputs, created_at, id, name, modified_at, source_thread_id, attachment_urls, error. Required: share_token.

create_a_lang_chain_examples_upload

Upload examples to a LangSmith dataset from a CSV file. Returns an array of created example objects including id, created_at, modified_at, name, dataset_id, inputs, and outputs. For non-CSV uploads, use the platform datasets examples endpoint instead. Required: upload_id.

create_a_lang_chain_examples_validate

Validate an example in LangSmith. Returns an ExampleValidationResult object with its id and validation attributes.

lang_chain_examples_validates_bulk_create

Validate examples in bulk in LangSmith. Returns an array of ExampleValidationResult objects, each with its id and validation attributes.

list_all_lang_chain_datasets

List LangSmith datasets with optional filters by name, data type, metadata, and tags. Returns an array of datasets including id, name, data_type, and created_at. Max 100 per page.

create_a_lang_chain_dataset

Create Dataset. Returns: name, description, created_at, inputs_schema_definition, outputs_schema_definition, externally_managed, transformations, data_type. Required: name.

lang_chain_datasets_bulk_delete

Delete multiple LangSmith datasets in a single request. Returns an empty 200 response on success. Max 100 dataset IDs per request. Required: dataset_ids.

update_a_lang_chain_dataset_by_id

Update a specific LangSmith dataset by id. Returns the updated dataset object including id, name, description, and timestamps. Required: id.

delete_a_lang_chain_dataset_by_id

Delete a specific dataset in LangSmith by id. Returns an empty response on success. Required: id.

lang_chain_datasets_bulk_update

Set a tag on a LangSmith dataset version identified by its as_of timestamp. Returns: tags, as_of. Required: dataset_id, as_of, tag.

get_single_lang_chain_dataset_by_id

Read Dataset. Returns: name, description, created_at, inputs_schema_definition, outputs_schema_definition, externally_managed, transformations, data_type. Required: id.

lang_chain_datasets_list_public

List publicly shared LangSmith datasets by share token. Returns: name, description, created_at, inputs_schema_definition, outputs_schema_definition, externally_managed, transformations, data_type, id, example_count, updated_at, last_session_start_time, metadata, thread_id, status, config. Required: share_token.

list_all_lang_chain_datasets_streams

Stream LangSmith datasets matching query parameters as a sequence of JSON patches. Returns streamed dataset records including id, name, and data_type. Max 100 per page.

create_a_lang_chain_datasets_upload

Upload a CSV or JSONL file to create a new dataset in LangSmith. Returns the created dataset object including id, name, data_type, and created_at.

create_a_lang_chain_datasets_upload_experiment

Upload an experiment that has already been run to LangSmith. Returns the experiment upload result object following the ExperimentResultsUploadResult schema. Required: experiment_name, results, experiment_start_time, experiment_end_time.

list_all_lang_chain_dataset_versions

List all versions of a LangSmith dataset, optionally filtered by a search string or a specific example. Returns: tags, as_of, tag, thread_id, created_at, updated_at, metadata, status, config. Max 100 per page. Required: dataset_id.

lang_chain_dataset_versions_get_version

Get a specific LangSmith dataset version by as_of timestamp or exact tag. Returns: tags, as_of, tag, thread_id, created_at, updated_at, metadata, status, config. Required: dataset_id.

list_all_lang_chain_versions_diffs

Get the diff between two dataset versions in LangSmith. Returns a DatasetDiffInfo object with version comparison attributes whose specific fields are not enumerated in the available API documentation. The from_version and to_version parameters each accept a date-time string or a version/tag identifier. Required: dataset_id, from_version, to_version.

list_all_lang_chain_dataset_openais

Download a LangSmith dataset in OpenAI Evals JSONL format. Returns an array of example records, each with content-type-specific attributes reflecting the dataset's inputs and outputs. Required: dataset_id.

list_all_lang_chain_dataset_openai_fts

Download a LangSmith dataset in OpenAI fine-tuning JSONL format. Returns records with messages containing role and content fields. Required: dataset_id.

list_all_lang_chain_dataset_csvs

Download a LangSmith dataset in CSV format. Returns the dataset as a CSV file download whose columns are determined by the dataset's example inputs and outputs. Optional: as_of to retrieve a specific point-in-time version. Required: dataset_id.

list_all_lang_chain_dataset_jsonls

Download a LangSmith dataset as a JSONL stream of example records. Returns: id, created_at, metadata, thread_id, updated_at, status, config. Required: dataset_id.

create_a_lang_chain_dataset_run

Fetch examples for a LangSmith dataset with associated runs filtered by session_ids. Returns: id, name, created_at, metadata. This endpoint is deprecated. Required: dataset_id, session_ids.

list_all_lang_chain_dataset_shares

Get the current share state of a LangSmith dataset. Returns: share_token, share_projects, thread_id, created_at, updated_at, metadata, status, config. The response is null when the dataset is not currently shared. Required: dataset_id.

lang_chain_dataset_shares_bulk_update

Share a LangSmith dataset, generating a public share token for read-only access via the public API. Returns: dataset_id, share_token, share_projects. Required: dataset_id.

lang_chain_dataset_shares_bulk_delete

Unshare a LangSmith dataset, revoking its public share token and removing public access. Returns an empty response on success. Required: dataset_id.

create_a_lang_chain_datasets_comparative

Create a comparative experiment in LangSmith. Returns: id, name, description, tenant_id, created_at, modified_at, reference_dataset_id, extra. Requires a JSON request body defining the experiment. Required: experiment_ids.

delete_a_lang_chain_datasets_comparative_by_id

Delete a specific comparative experiment in LangSmith by id. Returns an empty 204 response on success. Required: id.

list_all_lang_chain_datasets_comparatives

List shared comparative experiments for a dataset in LangSmith by share token. Returns: id, name, description, created_at, modified_at, extra, experiments_info, feedback_stats, thread_id, updated_at, metadata, status, config. Required: share_token.

create_a_lang_chain_datasets_clone

Clone a dataset in LangSmith. Returns an array of objects representing the clone results, with fields not specifically enumerated in the upstream OpenAPI spec. Required: target_dataset_id, source_dataset_id.

list_all_lang_chain_dataset_splits

List the split names for a LangSmith dataset, optionally as of a specific dataset version. Returns an array of split name strings. Required: dataset_id.

lang_chain_dataset_splits_bulk_update

Update dataset splits for a LangSmith dataset. Returns an array of UUID strings representing the updated split identifiers. Required: dataset_id, split_name, examples.

create_a_lang_chain_dataset_generate

Generate synthetic examples for a LangSmith dataset. Returns a JSON response with an unspecified schema. Required: dataset_id, num_examples.

create_a_lang_chain_datasets_studio_experiment

Run a LangSmith Studio experiment over a dataset by submitting a JSON request body conforming to the StudioRunOverDatasetRequestSchema. Returns a JSON object representing the experiment run result; field-level details for both the request and response are not enumerated in the available API reference. Requires a JSON request body. Required: project_name, dataset_id.

list_all_lang_chain_runs_rules

List all run rules in LangSmith with optional filters by dataset, session, type, name, evaluator, and tag values. Returns: id, tenant_id, is_enabled, session_id, session_name, dataset_id, dataset_name, display_name, sampling_rate, filter, trace_filter, tree_filter, add_to_annotation_queue_id, add_to_annotation_queue_name, add_to_dataset_id, add_to_dataset_name, add_to_dataset_prefer_correction,…

create_a_lang_chain_runs_rule

Create a new run rule in LangSmith. Returns: id, tenant_id, is_enabled, session_id, session_name, dataset_id, dataset_name, display_name, sampling_rate, filter, trace_filter, tree_filter, add_to_annotation_queue_id, add_to_annotation_queue_name, add_to_dataset_id, add_to_dataset_name, add_to_dataset_prefer_correction, corrections_dataset_id, use_corrections_dataset, num_few_shot_examples,…. Required: display_name, sampling_rate.

update_a_lang_chain_runs_rule_by_id

Update an existing run rule in LangSmith by id. Returns: id, tenant_id, is_enabled, session_id, session_name, dataset_id, dataset_name, display_name, sampling_rate, filter, trace_filter, tree_filter, add_to_annotation_queue_id, add_to_annotation_queue_name, add_to_dataset_id, add_to_dataset_name, add_to_dataset_prefer_correction, corrections_dataset_id, use_corrections_dataset,…. Required: id, display_name, sampling_rate.

delete_a_lang_chain_runs_rule_by_id

Delete a LangSmith run rule by id. Returns an empty 204 response on success. Required: id.

lang_chain_runs_rules_trigger

Manually trigger a LangSmith run rule by rule_id. Returns an empty 204 response on success. Required: rule_id.

lang_chain_runs_rules_list_logs

List logs for a particular LangSmith rule. Returns: rule_id, run_id, run_name, run_type, run_session_id, run_trace_id, start_time, end_time, application_time, add_to_annotation_queue, add_to_dataset, evaluators, alerts, webhooks, extend_only, thread_id, threads. Required: rule_id.

lang_chain_runs_rules_list_logs_v_2

List logs for a particular LangSmith rule with cursor-based pagination. Returns: logs, cursor. Max 1440 per page. Required: rule_id.

create_a_lang_chain_rules_validate

Validate a rule in LangSmith by executing it with test data without creating a saved rule. Returns an array of evaluation result objects in the same format as batch_invoke_evaluator. Only LLM-as-judge rules are supported; code evaluators are not allowed. Required: display_name, sampling_rate.

get_single_lang_chain_runs_thread_by_id

Get a preview of a thread in LangChain. Returns: thread_id, previews, created_at, updated_at, metadata, status, config, values. Required: id, session_id.

list_all_lang_chain_rule_last_applieds

Get Last Applied Rule. Returns: rule_id, run_id, run_name, run_type, run_session_id, run_trace_id, start_time, end_time. Required: rule_id.

create_a_lang_chain_rules_trigger

Trigger an array of run rules manually in LangSmith. Accepts a required JSON request body whose structure is defined by the TriggerRulesRequest schema. Returns an opaque success response with no defined fields.

get_single_lang_chain_run_by_id

Get a specific LangSmith run by id. Returns: name, inputs, inputs_preview, run_type, start_time, end_time, extra, error, execution_order, serialized, outputs, outputs_preview, parent_run_id, manifest_id, manifest_s3_id, events, tags, inputs_s3_urls, outputs_s3_urls, s3_urls, trace_id, dotted_order, trace_min_start_time, trace_max_start_time, id, status, child_run_ids, direct_child_run_ids,…. Required: id.

update_a_lang_chain_run_by_id

Update a LangSmith run by id with partial fields. Returns: message. Only the fields to be changed should be included in the body; unknown fields are ignored. Required: id.

create_a_lang_chain_run

Create (queue) a LangSmith run for ingestion. Returns: message.

lang_chain_runs_get_public

Get a shared LangSmith run by share token. Returns: name, inputs, inputs_preview, run_type, start_time, end_time, extra, error, execution_order, serialized, outputs, outputs_preview, parent_run_id, manifest_id, manifest_s3_id, events, tags, inputs_s3_urls, outputs_s3_urls, s3_urls, trace_id, dotted_order, trace_min_start_time, trace_max_start_time, id, status, child_run_ids, direct_child_run_ids,…. Required: share_token.

lang_chain_runs_get_public_by_id

Get a specific shared LangSmith run by share token and run id. Returns: name, inputs, inputs_preview, run_type, start_time, end_time, extra, error, execution_order, serialized, outputs, outputs_preview, parent_run_id, manifest_id, manifest_s3_id, events, tags, inputs_s3_urls, outputs_s3_urls, s3_urls, trace_id, dotted_order, trace_min_start_time, trace_max_start_time, id, status, child_run_ids,…. Required: run_id, share_token.

lang_chain_runs_get_public_v_2

Get a public shared trace run in LangSmith using a share token. Returns the run object including id, name, run_type, status, start_time, end_time, inputs, and outputs (returned fields depend on the selects parameter). Self-hosted deployments require LangSmith v0.16 or later. Required: share_token, run_id, start_time, selects.

lang_chain_runs_get_v_2

Get a single run in LangSmith by run ID for a given project. Returns the run object including id, name, run_type, status, start_time, end_time, inputs, and outputs (use selects to request specific fields; if omitted only id is returned). Self-hosted deployments require LangSmith v0.16 or later. Required: run_id, project_id.

list_all_lang_chain_run_shares

Get the sharing state of a LangSmith run. Returns the run's share_token, or null if the run is not currently shared. This endpoint is deprecated. Required: run_id.

create_a_lang_chain_run_share

Create or return a share token for a LangSmith run. Child runs share their trace root. Returns the share token response. Self-hosted deployments require LangSmith v0.16 or later. Required: run_id.

lang_chain_run_shares_delete_v_2

Delete the share token for a LangSmith run trace identified by trace_id and session_id. Idempotent. Returns an empty 204 response on success. Required: trace_id.

lang_chain_run_shares_bulk_update

Share a LangSmith run by enabling public sharing. Returns the run share object. This endpoint is deprecated. Required: run_id.

lang_chain_run_shares_bulk_delete

Unshare a LangSmith run by removing its share token. Returns an empty response on success. Deprecated endpoint. Required: run_id.

create_a_lang_chain_query_validate

Validate Runs Query. Returns: valid, errors.

list_all_lang_chain_runs_queries

Query Runs. Returns: runs, cursors, search_cursors, parsed_query.

lang_chain_runs_queries_list_public

Query Shared Runs. Returns: runs, cursors, parsed_query. Required: share_token.

lang_chain_runs_queries_list_public_dataset

Query shared dataset runs in LangSmith using a share token. Returns runs including id, trace_id, status, run_type, name, start_time, end_time, inputs, and outputs. Required: share_token.

lang_chain_runs_queries_list_public_v_2

Query all runs within a publicly shared LangSmith trace identified by a share token. Returns: items. Self-hosted deployments require LangSmith v0.16 or later. Required: share_token.

lang_chain_runs_queries_list_v_2

Query runs. Returns: items, next_cursor.

create_a_lang_chain_runs_generate_query

Generate a LangSmith runs filter expression from a natural language query. Returns: filter, feedback_urls. Required: query.

lang_chain_runs_generate_queries_create_public

Generate a runs filter expression from a natural language query for runs in projects executed over a shared LangSmith dataset. Returns the generated filter expression query. Required: share_token, query.

create_a_lang_chain_runs_stat

Get aggregated statistics for LangSmith runs by submitting a query in the request body. Returns a RunStats object with aggregate run metrics (or a map of RunStats objects keyed by group) computed over runs matching the query. Required: session.

lang_chain_runs_stats_create_public

Stats Shared Dataset Runs. Returns: run_count, latency_p50, latency_p99, first_token_p50, first_token_p99, total_tokens, prompt_tokens, completion_tokens. Required: share_token.

create_a_lang_chain_runs_batch

Ingest a batch of LangSmith runs via a single JSON payload containing post and/or patch arrays of run objects. Returns a 202 acknowledgment with per-run status messages (message). At least one of post or patch must be supplied. Prefer this endpoint for hundreds of runs; use /runs/multipart for very large fields or attachments.

create_a_lang_chain_runs_multipart

Ingest multiple runs, feedback objects, and binary attachments in a single multipart/form-data request to LangSmith. Returns a 202 Accepted response containing a map of part identifiers to status strings. Part names follow the pattern .[.] where event is post, patch, feedback, or attachment. Feedback parts must include trace_id and session_id. Each part must set a…

create_a_lang_chain_runs_group

Group Runs. Required: session_id, group_by.

create_a_lang_chain_group_stat

Stats Group Runs. Returns: run_count, latency_p50, latency_p99, first_token_p50, first_token_p99, total_tokens, prompt_tokens, completion_tokens. Required: session_id, group_by.

delete_a_lang_chain_delete_trace_by_id

Delete specific LangSmith runs by trace IDs within a project. Returns a 200 response on success; the response body is not documented by the API. Required: session_id, trace_ids.

delete_a_lang_chain_runs_delete_by_id

Delete LangSmith runs by trace ID or by time range. Supply session_id and trace_ids to delete specific traces, or start_time with optional end_time and metadata to delete matching traces. With both time bounds set, metadata is optional and omitting it deletes every trace in the range. Returns a 200 success response.

lang_chain_runs_deletes_delete_queue

Delete runs from a LangSmith annotation queue. This endpoint is deprecated. A request body is required but its properties are not documented in the available upstream source. Returns a 200 success response. Required: queue_id.

create_a_lang_chain_experiment_evaluate

Evaluate an existing experiment in LangSmith with a specific evaluator, triggering immediate batch evaluation of runs via the run_over_dataset approach. Returns an opaque JSON object with evaluation results whose fields are not enumerated in the source. A JSON request body specifying the evaluator configuration is required. Required: experiment_id, rule_id.

list_all_lang_chain_feedback_formulas

List feedback formulas for a given dataset or tracing project in LangSmith. Returns each formula's id and formula-specific attributes. Deprecated — superseded by composite-feedback v2.

create_a_lang_chain_feedback_formula

Create a new feedback formula in LangSmith. Returns the created feedback formula object including its id and formula-specific attributes. Deprecated — use POST /api/v1/feedback/composite-evaluators instead. Required: feedback_key, aggregation_type, formula_parts.

update_a_lang_chain_feedback_formula_by_id

Update a feedback formula in LangSmith by id. Returns the updated feedback formula object including its id and formula-specific attributes. Deprecated — superseded by composite-feedback v2. Required: id, feedback_key, aggregation_type, formula_parts.

delete_a_lang_chain_feedback_formula_by_id

Delete a LangSmith feedback formula by id. Returns an empty response on success. Deprecated: superseded by composite-feedback v2; tenants on v2 receive HTTP 410. Required: id.

get_single_lang_chain_feedback_formula_by_id

Get a LangSmith feedback formula by id. Returns the feedback formula object including its id and formula-specific attributes. Deprecated — superseded by composite-feedback v2. Required: id.

get_single_lang_chain_feedback_by_id

Get a specific LangSmith feedback by id. Returns: created_at, modified_at, key, score, value, comment, correction, feedback_group_id, comparative_experiment_id, run_id, session_id, id, trace_id, start_time, is_root, feedback_source, extra, feedback_thread_id, source, level, user_id. Required: id.

update_a_lang_chain_feedback_by_id

Replace an existing LangSmith feedback entry with a new, modified entry by id. Returns: created_at, modified_at, key, score, value, comment, correction, feedback_group_id, comparative_experiment_id, run_id, session_id, id, trace_id, start_time, is_root, feedback_source, extra, feedback_thread_id, source, level, user_id. Required: id.

delete_a_lang_chain_feedback_by_id

Delete a LangSmith feedback by id. Returns a 200 response with no documented response body on success. Required: id.

create_a_lang_chain_feedback

Create a new feedback entry in LangSmith. Returns the created feedback object including id, run_id, session_id, key, score, value, comment, correction, created_at, and modified_at. Required: key.

list_all_lang_chain_feedbacks

List feedback entries in LangSmith with optional filters by run, session, key, source, score presence, comment presence, level, and date range. Returns: created_at, modified_at, key, score, value, comment, correction, feedback_group_id, comparative_experiment_id, run_id, session_id, id, trace_id, start_time, is_root, feedback_source, extra, feedback_thread_id, level. Max 100 per page.

lang_chain_feedbacks_list_public

List feedback entries for a publicly shared LangSmith project by share token. Returns: created_at, modified_at, key, score, value, comment, correction, feedback_group_id, comparative_experiment_id, run_id, session_id, id, trace_id, start_time, is_root, feedback_source, extra, feedback_thread_id, source, level, user_id. Max 100 per page. Required: share_token.

list_all_lang_chain_feedback_tokens

List all feedback ingest tokens for a run in LangSmith. Returns: id, url, expires_at, feedback_key, run_id, created_at. Required: run_id.

create_a_lang_chain_feedback_token

Create a new feedback ingest token in LangSmith. Returns the created token including id, created_at, and webhooks configuration.

get_single_lang_chain_feedback_token_by_id

Create new feedback for a run using a feedback ingest token in LangSmith via GET. Accepts optional score, value, comment, correction, and extend_trace_retention query parameters. The API returns a 200 response with no specifically documented response fields. Required: id.

lang_chain_feedback_tokens_attach

Create new feedback for a run using a feedback ingest token in LangSmith. Accepts webhook configuration in the request body. Returns the feedback result including webhooks configuration. Required: token_id.

list_all_lang_chain_datasets_sessions

List tracer sessions (projects) run on a shared LangSmith dataset using a share token. Returns: start_time, end_time, extra, name, description, default_dataset_id, reference_dataset_id, trace_tier, id, run_count, latency_p50, latency_p99, first_token_p50, first_token_p99, total_tokens, prompt_tokens, completion_tokens, total_cost, prompt_cost, completion_cost, tenant_id, last_run_start_time,…. Required: share_token.

list_all_lang_chain_datasets_sessions_bulks

List tracer sessions from multiple shared LangSmith datasets in a single bulk request. Returns: start_time, end_time, extra, name, description, default_dataset_id, reference_dataset_id, trace_tier, id, run_count, latency_p50, latency_p99, first_token_p50, first_token_p99, total_tokens, prompt_tokens, completion_tokens, total_cost, prompt_cost, completion_cost, tenant_id, last_run_start_time,…. Required: share_tokens.

get_single_lang_chain_datasets_run_by_id

Get a single run from a shared LangSmith dataset by run id. Returns: name, inputs, inputs_preview, run_type, start_time, end_time, extra, error, execution_order, serialized, outputs, outputs_preview, parent_run_id, manifest_id, manifest_s3_id, events, tags, inputs_s3_urls, outputs_s3_urls, s3_urls, trace_id, dotted_order, trace_min_start_time, trace_max_start_time, id, status, child_run_ids,…. Required: id, share_token.

list_all_lang_chain_datasets_feedbacks

List feedback for runs in projects executed over a shared dataset in LangSmith. Returns: created_at, modified_at, key, score, value, comment, correction, feedback_group_id, comparative_experiment_id, run_id, session_id, id, trace_id, start_time, is_root, feedback_source, extra, feedback_thread_id, source, level, thread_id, updated_at, metadata, status, config. Required: share_token.

list_all_lang_chain_schema_messages

Get the JSON Schema definition for messages in LangSmith at a specified version. Returns a version-specific JSON Schema document describing the message format. Required: version.

list_all_lang_chain_schema_tooldefs

Get the LangSmith tool definition JSON Schema for a specific schema version. Returns the JSON Schema document that defines the tool definition format; the exact fields are version-specific. Required: version.

list_all_lang_chain_annotation_queues

List annotation queues in LangSmith with optional filters by name, queue type, dataset, tags, and assignment. Returns: description, created_at, updated_at, num_reviewers_per_item, enable_reservations, reservation_minutes, reviewer_access_mode, name, id, tenant_id, source_rule_id, run_rule_id, default_dataset, queue_type, metadata, assigned_reviewers, total_runs. Max 100 per page.

create_a_lang_chain_annotation_queue

Create a new annotation queue in LangSmith. Returns: description, created_at, updated_at, num_reviewers_per_item, enable_reservations, reservation_minutes, reviewer_access_mode, name, id, tenant_id, source_rule_id, run_rule_id, default_dataset, queue_type, metadata, assigned_reviewers. Required: name.

lang_chain_annotation_queues_bulk_delete

Delete multiple annotation queues in LangSmith with partial success support. Returns 200 when all queues are deleted successfully, or 207 for partial success. Max 100 queue IDs per request. Required: queue_ids.

update_a_lang_chain_annotation_queue_by_id

Update a LangSmith annotation queue by id. Returns: id, name, queue_type. Required: id.

delete_a_lang_chain_annotation_queue_by_id

Delete a LangSmith annotation queue by id. Returns an empty 204 response on success. Required: id.

get_single_lang_chain_annotation_queue_by_id

Get a single LangSmith annotation queue by id. Returns: description, created_at, updated_at, num_reviewers_per_item, enable_reservations, reservation_minutes, reviewer_access_mode, name, id, tenant_id, source_rule_id, run_rule_id, default_dataset, queue_type, metadata, assigned_reviewers, rubric_items, rubric_instructions, dataset_id. Required: id.

create_a_lang_chain_annotation_queues_populate

Populate a LangSmith annotation queue with runs from an experiment. Accepts a required JSON request body (PopulateAnnotationQueueSchema) whose field definitions are not available in the discovered API documentation. Returns a success response whose body shape is not specified by the API. Required: queue_id, session_ids.

list_all_lang_chain_annotation_queue_runs

List runs from a LangSmith annotation queue, filterable by archived state, stats inclusion, and review status. Returns: name, inputs, inputs_preview, run_type, start_time, end_time, extra, error, execution_order, serialized, outputs, outputs_preview, parent_run_id, manifest_id, manifest_s3_id, events, tags, inputs_s3_urls, outputs_s3_urls, s3_urls, trace_id, dotted_order, trace_min_start_time,…. Required: queue_id.

create_a_lang_chain_annotation_queue_run

Add runs to a LangSmith annotation queue by providing an array of run UUIDs or run objects. Returns an array of annotation queue run objects including id and name. This endpoint is deprecated. Required: queue_id.

update_a_lang_chain_annotation_queue_run_by_id

Update a run in a LangSmith annotation queue, typically to change its review status. Returns: id, name. This endpoint is deprecated. Required: id, queue_id.

delete_a_lang_chain_annotation_queue_run_by_id

Delete a run from a LangSmith annotation queue. Returns an empty 204 response on success. This endpoint is deprecated. Required: id, queue_id.

lang_chain_annotation_queue_runs_get_by_index

Get a single run from a LangSmith annotation queue by its zero-based index position. Returns: name, inputs, inputs_preview, run_type, start_time, end_time, extra, error, execution_order, serialized, outputs, outputs_preview, parent_run_id, manifest_id, manifest_s3_id, events, tags, inputs_s3_urls, outputs_s3_urls, s3_urls, trace_id, dotted_order, trace_min_start_time, trace_max_start_time, id,…. Required: queue_id, index.

create_a_lang_chain_runs_by_key

Add runs to a LangSmith annotation queue by key. Accepts an array of run key request objects as the request body. Returns an array of annotation queue run objects including id. This endpoint is deprecated; self-hosted deployments require LangSmith v0.16 or later. Required: queue_id.

create_a_lang_chain_annotation_queue_export

Export archived runs from a LangSmith annotation queue. The response payload has an unspecified schema; no response fields are documented upstream. Required: queue_id.

list_all_lang_chain_annotation_queue_queues

List annotation queues associated with a specific run in LangSmith. Returns: description, created_at, updated_at, num_reviewers_per_item, enable_reservations, reservation_minutes, reviewer_access_mode, name, id, tenant_id, source_rule_id, run_rule_id, default_dataset, queue_type, metadata, assigned_reviewers. Required: run_id.

list_all_lang_chain_annotation_queue_total_sizes

Get the total size (item count) of a LangSmith annotation queue across all review statuses. Returns: size. This endpoint is deprecated. Required: queue_id.

list_all_lang_chain_annotation_queue_total_archiveds

Get the total count of archived items from a LangSmith annotation queue, optionally filtered by a time range. Returns: size. Optional: start_time, end_time. This endpoint is deprecated. Required: queue_id.

list_all_lang_chain_annotation_queue_sizes

Get the size of a LangSmith annotation queue, optionally filtered by review status. Returns: size. This endpoint is deprecated. Required: queue_id.

create_a_lang_chain_annotation_queues_status

Create an identity annotation queue run status in LangSmith. Returns an unspecified response body whose schema is not documented by the upstream API. This endpoint is deprecated. Required: statu_id.

get_single_lang_chain_runs_resolve_by_id

Resolve a LangSmith annotation queue run ID to its section and run data for deep linking. Returns: section, position. This endpoint is deprecated. Required: id, queue_id.

create_a_lang_chain_ace_execute

Execute custom code for testing purposes in LangSmith. Returns an opaque JSON object whose shape depends on the code executed. A JSON request body is required. Required: args, code, language.

list_all_lang_chain_bulk_exports

List all bulk exports in the current LangSmith workspace. Returns each export with id, status, start_time, bulk_export_destination_id, and source_bulk_export_id.

create_a_lang_chain_bulk_export

Create a new LangSmith bulk export to export tracing runs to a destination. Returns: bulk_export_destination_id, session_id, all_experiments, start_time, end_time, filter, format, format_version, compression, interval_hours, export_fields, id, tenant_id, status, created_at, updated_at, finished_at, source_bulk_export_id. Either session_id or…. Required: bulk_export_destination_id, start_time.

update_a_lang_chain_bulk_export_by_id

Cancel a LangSmith bulk export by id. Returns the updated bulk export including id, status, and source_bulk_export_id. Required: id.

get_single_lang_chain_bulk_export_by_id

Get a single LangSmith bulk export by id. Returns: bulk_export_destination_id, session_id, all_experiments, start_time, end_time, filter, format, format_version, compression, interval_hours, export_fields, id, tenant_id, status, created_at, updated_at, finished_at, source_bulk_export_id. Required: id.

list_all_lang_chain_bulk_exports_destinations

List all bulk export destinations configured in the current LangSmith workspace. Returns: destination_type, display_name, config, id, tenant_id, created_at, updated_at, credentials_keys, webhooks.

create_a_lang_chain_bulk_exports_destination

Create a new bulk export destination in LangSmith that writes exported data to an S3 or S3-compatible bucket. Credentials are stored encrypted and the API validates the destination before saving. Returns: destination_type, display_name, config, id, tenant_id, created_at, updated_at, credentials_keys. Required: display_name, config.

update_a_lang_chain_bulk_exports_destination_by_id

Update a LangSmith bulk export destination by id. Returns the updated destination object including id, destination_type, display_name, config, and credentials. Required: id.

get_single_lang_chain_bulk_exports_destination_by_id

Get a single LangSmith bulk export destination by id. Returns: destination_type, display_name, config, id, tenant_id, created_at, updated_at, credentials_keys, credentials. Required: id.

list_all_lang_chain_bulk_exports_runs

List LangSmith bulk export runs spawned by a scheduled bulk export, filtered by the source export's id. Returns: bulk_export_id, metadata, session_id, id, status, retry_number, errors, created_at, updated_at, finished_at, start_time, end_time, source_bulk_export_id, tenant_id. Required: source_bulk_export_id.

list_all_lang_chain_bulk_export_runs

List runs for a LangSmith bulk export. Returns an array of BulkExportRun objects including id, tenant_id, session_id, trace_id, name, run_type, status, start_time, end_time, inputs, and outputs. Max 1000 per page. Required: bulk_export_id.

get_single_lang_chain_bulk_export_run_by_id

Get a single LangSmith bulk export run by id. Returns the full BulkExportRun object including id, tenant_id, session_id, trace_id, name, run_type, status, start_time, end_time, inputs, and outputs. Required: id, bulk_export_id.

list_all_lang_chain_feedback_configs

List LangSmith feedback configs with optional filtering by feedback key and name. Returns: feedback_key, feedback_config, tenant_id, modified_at, is_lower_score_better.

create_a_lang_chain_feedback_config

Create a new LangSmith feedback config for a given feedback key. Returns: feedback_key, feedback_config, tenant_id, modified_at, is_lower_score_better. Required: feedback_key, feedback_config.

lang_chain_feedback_configs_bulk_update

Update an existing LangSmith feedback config identified by its feedback key. Returns: feedback_key, feedback_config, tenant_id, modified_at, is_lower_score_better. Required: feedback_key.

lang_chain_feedback_configs_bulk_delete

Soft delete a LangSmith feedback config by marking it as deleted. The config can be recreated later with the same key, and existing feedback records with this key remain unchanged. Returns an empty 204 response on success. Required: feedback_key.

list_all_lang_chain_model_price_maps

List model price map entries in LangSmith. Returns an array of model price map records including id, created_at, and updated_at. The optional q parameter filters results by a search query string (max 200 characters).

create_a_lang_chain_model_price_map

Create a new model price entry in LangSmith. The request body schema (ModelPriceMapCreateSchema) and response shape are defined upstream but not enumerated in the API docs. Required: name, match_pattern, prompt_cost, completion_cost.

update_a_lang_chain_model_price_map_by_id

Update a model price map entry in LangSmith by id. The request body schema (ModelPriceMapUpdateSchema) and response shape are defined upstream but not enumerated in the API docs. Required: id, name, match_pattern, prompt_cost, completion_cost.

delete_a_lang_chain_model_price_map_by_id

Delete a model price map entry in LangSmith by id. Returns the deleted record including its id and attributes. Required: id.

create_a_lang_chain_prompts_invoke_prompt

Invoke a prompt in LangSmith by submitting a JSON request body conforming to the InvokePromptPayload schema. Returns a dynamic JSON response whose shape depends on the invoked prompt. Required: messages, template_format, inputs.

list_all_lang_chain_prompt_webhooks

List all prompt webhooks for the current tenant in LangSmith. Returns each webhook object including id and configuration attributes.

create_a_lang_chain_prompt_webhook

Create a new prompt webhook in LangSmith. Returns the created webhook object including id and configuration attributes. A JSON request body is required. Required: url.

update_a_lang_chain_prompt_webhook_by_id

Update a specific prompt webhook in LangSmith by id. Returns the updated webhook object including id and configuration attributes. A JSON request body is required. Required: id.

delete_a_lang_chain_prompt_webhook_by_id

Delete a specific prompt webhook in LangSmith by id. Returns an empty 204 response on success. Required: id.

get_single_lang_chain_prompt_webhook_by_id

Get a specific prompt webhook in LangSmith by id. Returns the webhook object including id and configuration attributes. Required: id.

create_a_lang_chain_prompt_webhooks_test

Test a specific prompt webhook in LangSmith by sending a test request to its configured target. Returns a free-form object with string key-value pairs representing the webhook test response. Requires a PromptWebhookTest request body whose individual fields are not enumerated in the API reference. Required: webhook, payload.

list_all_lang_chain_playground_settings

List all playground settings for the current tenant in LangSmith. Returns: id, settings, options, name, organization_id, created_at, updated_at, description, settings_type, created_by_ls_user_id, updated_by_ls_user_id, available_in_playground, available_in_evaluators, available_in_agent_builder, available_in_polly, available_in_insights_heavy, available_in_insights_light, oauth_enabled,…

create_a_lang_chain_playground_setting

Create playground settings in LangSmith. Returns the created record including id, created_at, and updated_at. Requires a JSON request body. Required: settings.

update_a_lang_chain_playground_setting_by_id

Update an existing playground settings record in LangSmith. Returns the updated record including id, created_at, and updated_at. Required: id.

delete_a_lang_chain_playground_setting_by_id

Delete a playground settings record in LangSmith by id. Returns an empty 200 response on success. Required: id.

get_single_lang_chain_playground_setting_by_id

Get a single playground settings record in LangSmith by id. Returns: id, settings, options, name, organization_id, created_at, updated_at, description, settings_type, created_by_ls_user_id, updated_by_ls_user_id, available_in_playground, available_in_evaluators, available_in_agent_builder, available_in_polly, available_in_insights_heavy, available_in_insights_light, oauth_enabled,…. Required: id.

create_a_lang_chain_section_clone

Clone a dashboard section in LangSmith. Returns the cloned section object including its id and section-specific attributes. A JSON request body is required.

list_all_lang_chain_charts_sections

List all charts sections (dashboard sections) for the tenant in LangSmith. Returns: title, description, index, id, chart_count, created_at, modified_at. Max 100 per page.

create_a_lang_chain_charts_section

Create a new charts section in LangSmith. Returns the created section object. The request body and response schema are referenced as CustomChartsSectionCreate and CustomChartsSectionResponse in the upstream API; their individual fields are not enumerated in the discovered source. Required: title.

update_a_lang_chain_charts_section_by_id

Update a charts section by id in LangSmith. Returns the updated section object. The request body and response schema are referenced as CustomChartsSectionUpdate and CustomChartsSectionResponse in the upstream API; their individual fields are not enumerated in the discovered source. Required: id.

delete_a_lang_chain_charts_section_by_id

Delete a LangSmith charts section by id. Returns an empty response on success. Required: id.

get_single_lang_chain_charts_section_by_id

Get a single charts section by id in LangSmith. Returns the section object. The request body and response schema are referenced as CustomChartsSectionRequest and CustomChartsSection in the upstream API; their individual fields are not enumerated in the discovered source. Required: id.

update_a_lang_chain_chart_by_id

Update a chart or text block in LangSmith by ID. Returns: id. The response uses a chart_type discriminator with values line, bar, table, kpi, top-k, pie, or text. Required: id.

delete_a_lang_chain_chart_by_id

Delete a LangSmith chart by id. Returns an empty response on success. Required: id.

list_all_lang_chain_charts

List all custom charts and dashboard text blocks for a LangSmith tenant. Returns: sections, webhooks. The response uses a chart_type discriminator with values line, bar, table, kpi, top-k, pie, or text.

get_single_lang_chain_chart_by_id

Get a single custom chart or text block by ID in LangSmith. Returns: id, webhooks. The response uses a chart_type discriminator with values line, bar, table, kpi, top-k, pie, or text. Required: id.

create_a_lang_chain_charts_preview

Read Chart Preview. Returns: series_id, timestamp, value, group. Required: bucket_info, chart.

create_a_lang_chain_charts_create

Create a chart or dashboard text block in LangSmith. The request body is discriminated by chart_type: line, bar, table, kpi, top-k, and pie create a chart; text creates a text block. Returns: chart_type.

list_all_lang_chain_org_charts_sections

List organization chart sections in LangSmith. Deprecated — organization usage charts have been retired and the endpoint returns HTTP 410 Gone.

create_a_lang_chain_org_charts_section

Create an organization chart section in LangSmith. Deprecated — organization usage charts have been retired and the endpoint returns HTTP 410 Gone.

update_a_lang_chain_org_charts_section_by_id

Update an organization chart section by id in LangSmith. Deprecated — organization usage charts have been retired and the endpoint returns HTTP 410 Gone. Required: id.

delete_a_lang_chain_org_charts_section_by_id

Delete an organization chart section by id in LangSmith. Deprecated — organization usage charts have been retired and the endpoint returns HTTP 410 Gone. Required: id.

get_single_lang_chain_org_charts_section_by_id

Get a single organization chart section by id in LangSmith. Deprecated — organization usage charts have been retired and the endpoint returns HTTP 410 Gone. Required: id.

update_a_lang_chain_org_chart_by_id

Update an organization usage chart in LangSmith. This endpoint is deprecated — organization usage charts have been retired and the upstream returns HTTP 410. The source documents no request body fields or response schema for this deprecated endpoint. Required: id.

delete_a_lang_chain_org_chart_by_id

Delete an organization usage chart in LangSmith. This endpoint is deprecated — organization usage charts have been retired and the upstream returns HTTP 410. Returns an empty response on success. Required: id.

list_all_lang_chain_org_charts

List organization usage charts in LangSmith. This endpoint is deprecated — organization usage charts have been retired and the upstream returns HTTP 410. The source documents no request body fields or response schema for this deprecated endpoint.

get_single_lang_chain_org_chart_by_id

Get a single organization usage chart in LangSmith. This endpoint is deprecated — organization usage charts have been retired and the upstream returns HTTP 410. The source documents no request body fields or response schema for this deprecated endpoint. Required: id.

create_a_lang_chain_org_charts_preview

Preview an organization usage chart in LangSmith. This endpoint is deprecated — organization usage charts have been retired and the upstream API returns HTTP 410 Gone. No request body or query parameters are accepted. Returns an empty response on success.

create_a_lang_chain_org_charts_create

Create an organization usage chart in LangSmith. This endpoint is deprecated — organization usage charts have been retired and it returns HTTP 410 Gone with no response body.

list_all_lang_chain_mcp_tools

List MCP tools from a remote server URL in LangSmith, served from cache when fresh and fetched via manifest or MCP handshake on cache miss. Returns the tools exposed by the remote MCP server; response shape is server-specific. Required: url.

lang_chain_mcp_tools_bulk_delete

Invalidate cached MCP tools for a given server URL in LangSmith so subsequent requests re-fetch from the remote server. Returns a successful response with no documented body. Required: url.

list_all_lang_chain_mcp_proxies

Proxy a GET request to a remote MCP server URL through LangSmith. Returns the proxied response from the target server, whose shape depends on the requested URL. Required: url.

create_a_lang_chain_mcp_proxy

Proxy a request to a remote MCP server through LangSmith by POSTing a ProxyRequest body. Returns the proxied response from the target server, whose shape is dynamic and depends on the requested URL and payload. Required: url.

list_all_lang_chain_orgs

List all LangSmith organizations visible to the current authentication. Returns: id, display_name, tier, created_at, created_by_user_id, created_by_ls_user_id, modified_at, is_personal, disabled, sso_login_slug, sso_only, jit_provisioning_enabled, invites_enabled, public_sharing_disabled, pat_creation_disabled, workspace_admin_can_invite_to_org, byoc_create_saas_workspace_enabled,…

create_a_lang_chain_org

Create a new organization in LangSmith. Returns: id, display_name, tier, created_at, created_by_user_id, created_by_ls_user_id, modified_at, is_personal, disabled, sso_login_slug, sso_only, jit_provisioning_enabled, invites_enabled, public_sharing_disabled, pat_creation_disabled, workspace_admin_can_invite_to_org, byoc_create_saas_workspace_enabled, default_sso_provision, max_api_key_expiry_days,…. Required: display_name, is_personal.

get_single_lang_chain_org_by_id

Get Organization Info. Returns: id, display_name, config, connected_to_stripe, connected_to_metronome, is_personal, tier, payment_method.

create_a_lang_chain_orgs_setup

Create a Stripe customer and retrieve a Stripe SetupIntent for the current LangSmith organization's billing setup. Returns: client_secret. No required parameters.

list_all_lang_chain_orgs_info

Get current organization info in LangSmith. Returns the OrganizationInfo object for the authenticated user's current organization, including its id, name, and configuration details.

get_single_lang_chain_orgs_info_by_id

Get current organization info. Returns: byoc_create_saas_workspace_enabled, can_export_usage_backfill, config, default_sso_provision, disabled, disabled_model_providers, display_name, engine_enabled.

lang_chain_orgs_info_bulk_update

Update Current Organization Info. Returns: id, display_name, config, engine_enabled, engine_lcu_spend_limit_monthly, is_personal, tier, reached_max_workspaces.

list_all_lang_chain_orgs_billings

Get the current organization's billing information in LangSmith. Returns: id, display_name, config, connected_to_stripe, connected_to_metronome, is_personal, tier, payment_method, end_of_billing_period, current_plan, upcoming_plan, reached_max_workspaces, disabled, default_sso_provision, plus_plan_transition.

list_all_lang_chain_orgs_dashboards

Get the current organization's dashboard in LangSmith filtered by dashboard type and color scheme. Returns the dashboard object with type-specific attributes. Required: type, color_scheme.

create_a_lang_chain_orgs_payment_method

Notify LangSmith that a payment method was created by submitting Stripe payment information. Returns a 200 acknowledgment response with an unspecified JSON body. Required: billing_info, setup_intent.

list_all_lang_chain_orgs_business_infos

Get the current organization's company info from LangSmith. Returns the company/business info object (StripeBusinessInfo-Output). The specific fields of the returned object are defined by the upstream schema and were not enumerated in the discovered source.

create_a_lang_chain_orgs_business_info

Set the current organization's company/business info in LangSmith. Accepts a required JSON request body whose fields are defined by the upstream StripeBusinessInfo-Input schema, which is not enumerated in the discovered source. Returns a 200 success response whose body shape is not defined in the API specification.

create_a_lang_chain_orgs_plan

Change the payment plan for the current LangSmith organization. Accepts a request body conforming to the ChangePaymentPlanSchema whose field definitions are not enumerated in the discovered source. Returns a 200 response with an unspecified JSON body. Required: tier.

list_all_lang_chain_orgs_roles

List Organization Roles. Returns: id, name, display_name, description, organization_id, permissions, access_scope, is_restricted.

create_a_lang_chain_orgs_role

Create a new organization role in LangSmith. Returns the created role object including its id and attributes. A JSON request body with the role definition is required. Required: display_name, description, permissions.

update_a_lang_chain_orgs_role_by_id

Update a LangSmith organization role by id. Returns the updated role object including its id and attributes. Required: id, display_name, description, permissions.

delete_a_lang_chain_orgs_role_by_id

Delete a LangSmith organization role by id. Returns the deleted role object including its id and attributes. Required: id.

lang_chain_role_restrictions_bulk_update

Set the restriction for a LangSmith organization role by role_id. Returns the updated Role object; the upstream API reference does not enumerate its fields in the discovered source, so consult the LangSmith Smith API reference for the full field-level breakdown. Required: role_id, is_restricted.

list_all_lang_chain_orgs_permissions

List all available organization permissions in LangSmith. Returns: name, description, access_scope.

list_all_lang_chain_orgs_pendings

List all pending LangSmith organization invites visible to the current auth. Returns an array of organization objects including id, name, and created_at.

delete_a_lang_chain_orgs_pending_by_id

Delete a pending organization invite in LangSmith by id. Returns an empty 204 response on success. Required: id.

list_all_lang_chain_orgs_members

List members of the current LangSmith organization. Returns: organization_id, members, pending, id, display_name, email, organization_role, workspace_role, workspace_memberships, active, created_at.

create_a_lang_chain_orgs_member

Add a member to the current LangSmith organization by creating a pending identity invitation. Returns the created pending identity including id. The request body uses the PendingIdentityCreate schema whose properties are not enumerated in the available source. Required: email.

update_a_lang_chain_orgs_member_by_id

Update a LangSmith organization member's role (all auth modes) or full_name/password (basic auth). Returns: id. Required: id.

delete_a_lang_chain_orgs_member_by_id

Remove a member from the current LangSmith organization. Returns an empty response on success. Required: id.

get_single_lang_chain_orgs_member_by_id

List members of the current LangSmith organization (active and pending) enriched with workspace memberships. Returns: avatar_url, created_at, display_name, email, id, is_disabled, is_pending, linked_login_methods, ls_user_id, role_id, role_name, scim_groups, workspace_memberships. Optional filters include active_is, pending_is, name_like, email_like, workspace_name_like, organization_role_like,…

list_all_lang_chain_members_actives

List active members of the current organization in LangSmith. Returns an array of member identity objects including id, email, name, created_at, and is_disabled. The user_ids filter is deprecated; use ls_user_ids instead.

get_single_lang_chain_members_active_by_id

List active members of the current workspace in LangSmith. Returns an array of member identity objects including id, email, name, created_at, and is_disabled. The user_ids filter is deprecated; use ls_user_ids instead.

list_all_lang_chain_members_pendings

List pending org member invites for the current LangSmith organization, with optional filtering by emails and search by email or name. Returns an array of pending identities with id, email, name, and created_at. Max 500 per page.

get_single_lang_chain_members_pending_by_id

Get pending member invitations for the current LangSmith workspace. Returns an array of pending identities with id, email, name, and created_at. Supports filtering by emails and searching by email or name. Max 500 results per page.

create_a_lang_chain_members_batch

Add Members To Current Org Batch. Returns: email, read_only, role_id, role_name, workspace_ids, workspace_role_id, workspace_role_name, password.

lang_chain_members_batches_create_workspace

Batch invite up to 500 users to the current LangSmith workspace and organization. Returns an array of invite results, one per requested user. Rate limited per workspace to reduce bulk invitation abuse.

create_a_lang_chain_basic_batch

Add Basic Auth Members To Current Org. Returns: id, ls_user_id, created_at, updated_at, email, full_name, avatar_url, password.

lang_chain_member_pendings_update_workspace

Update the role on a pending workspace member invite in LangSmith. Returns the updated pending member identity including identity_id and role. Required: identity_id.

lang_chain_member_pendings_delete_workspace

Delete a pending workspace member invite in LangSmith. Returns an empty response on success. Required: identity_id.

lang_chain_member_pendings_bulk_update

Update the role on a pending LangSmith organization member invite. Returns: email, read_only, role_id, role_name, workspace_ids, workspace_role_id, workspace_role_name, password, full_name, access_scope, id, user_id, tenant_id, organization_id, created_at, org_role_id, org_role_name, tenant_ids, role. Required: identity_id.

lang_chain_member_pendings_bulk_delete

Delete a pending LangSmith organization member invite by identity_id. Returns an empty response on success. Required: identity_id.

create_a_lang_chain_claim

Claim a pending organization invite in LangSmith by accepting the invitation for the authenticated user. Returns the Identity object including id. Required: organization_id.

lang_chain_claims_claim_workspace

Claim a pending workspace invite in LangSmith by accepting the invitation for the authenticated user. This endpoint is deprecated. Returns an opaque response with no documented fields. Required: workspace_id.

lang_chain_members_basics_bulk_update

Update the current LangSmith user's full_name and/or password in basic auth mode. The API returns a success response whose body shape is not specified in the API reference.

list_all_lang_chain_orgs_sso_settings

Get Current Sso Settings. Returns: id, organization_id, provider_id, default_workspace_role_id, default_workspace_ids, metadata_url, metadata_xml, sso_groups_enabled.

create_a_lang_chain_orgs_sso_setting

Create SSO provider settings for the current organization in LangSmith. Returns the created SSO provider object with its id and provider-specific attributes. Requires a JSON request body with SSO provider configuration. Required: default_workspace_role_id, default_workspace_ids.

update_a_lang_chain_orgs_sso_setting_by_id

Update SSO provider settings defaults for the current organization in LangSmith. Returns the updated SSO provider record including its id and sso_login_slug. Required: id.

delete_a_lang_chain_orgs_sso_setting_by_id

Delete Sso Settings. Returns: id, organization_id, provider_id, default_workspace_role_id, default_workspace_ids, metadata_url, metadata_xml, sso_groups_enabled. Required: id.

lang_chain_orgs_login_methods_bulk_update

Update allowed login methods for the current organization in LangSmith. Returns an object representing the updated allowed login methods configuration. Requires a request body following the AllowedLoginMethodsUpdate schema.

list_all_lang_chain_billing_usages

List organization billing usage in LangSmith for a given date range. Returns an array of OrgUsage records with trace consumption data for the specified period. Required: starting_on.

list_all_lang_chain_billing_granular_usages

List granular LangSmith usage data for the current organization with flexible grouping by workspace or trace tier. Returns time-bucketed records with traces, nodes_executed, agent_runs, and agent_uptime_seconds, with value fields varying by kind. Required: start_time, end_time.

list_all_lang_chain_granular_usage_exports

Export granular usage data as a CSV file in LangSmith. Returns a CSV file download whose value columns vary by the kind parameter. Required: start_time, end_time.

list_all_lang_chain_user_login_methods

Get Current User Login Methods. Returns: id, provider, ls_user_id, saml_provider_id, provider_user_id, created_at, updated_at, email.

create_a_lang_chain_orgs_stripe_checkout_session

Create a Stripe checkout session in LangSmith to kick off a checkout flow for the current organization. Returns the Stripe checkout session object. Requires a JSON request body whose field schema is not enumerated in the available API documentation. Required: amount_cents, success_path.

create_a_lang_chain_orgs_stripe_account_link

Kick off a Stripe account link flow in LangSmith. Requires a JSON request body (StripeAccountLinksCreate schema). Returns a JSON response whose field structure is not enumerated in the available source documentation. Required: success_path.

list_all_lang_chain_orgs_service_keys

List LangSmith org service keys for the current organization, optionally filtered by workspace. Returns an array of service key objects including id, description, and read_only for each key. Optional: workspace_ids.

create_a_lang_chain_orgs_service_key

Create an org-scoped service key in LangSmith. If workspaces is None, the key is org-wide. Returns: created_at, id, short_key, description, read_only, last_used_at, expires_at, workspace_names, default_workspace_name, role_id, org_role_id, access_scope, created_by, key.

update_a_lang_chain_orgs_service_key_by_id

Update a LangSmith org service key's role(s) in place without rotating the key. Returns: created_at, id, short_key, description, read_only, last_used_at, expires_at, workspace_names, default_workspace_name, role_id, org_role_id, access_scope, created_by, key. Required: id.

delete_a_lang_chain_orgs_service_key_by_id

Delete a LangSmith org service key by id. Returns: created_at, id, short_key, description, read_only, last_used_at, expires_at, workspace_names, default_workspace_name, role_id, org_role_id, access_scope, created_by, key. Required: id.

list_all_lang_chain_orgs_personal_access_tokens

List personal access tokens (PATs) for the current LangSmith organization. Returns each token with id, description, read_only, and created_at.

create_a_lang_chain_orgs_personal_access_token

Create a personal access token (PAT) for the current LangSmith organization. The API key value is displayed only once in the response. Returns: created_at, id, short_key, description, read_only, last_used_at, expires_at, workspace_names, default_workspace_name, role_id, org_role_id, access_scope, created_by, key, api_key.

delete_a_lang_chain_orgs_personal_access_token_by_id

Delete a LangSmith personal access token for the current organization by id. Returns the deleted token object including its id, description, and read_only flag. Required: id.

create_a_lang_chain_orgs_set_default_sso_provision

Set the current LangSmith organization as the default for SSO provisioning in self-hosted environments. Returns a success response object. No request parameters required.

create_a_lang_chain_login

Create a login session in LangSmith by authenticating with credentials. Returns a BasicAuthResponse containing authentication details.

create_a_lang_chain_email_verification_send

Send an email to confirm the email address for an SSO user in LangSmith. Returns a 202 acknowledgment object with an opaque body. Requires a JSON request body with SSO email verification send details. Required: email, saml_provider_id.

create_a_lang_chain_email_verification_status

Check the email verification status of an SSO user in LangSmith. Returns the SSO email verification status response object. Required: email, saml_provider_id.

create_a_lang_chain_email_verification_confirm

Confirm the email of an SSO user in LangSmith by submitting a JSON verification request. Returns a generic confirmation response object whose specific fields are not enumerated in the available source. The request requires a JSON body (SSOConfirmEmailRequest). Required: token.

get_single_lang_chain_sso_setting_by_id

Get SSO provider settings from a login slug in LangSmith. Returns an array of SSO provider settings records (SSOProviderSlim); the specific fields of each record are not enumerated in the available API reference. Required: id.

create_a_lang_chain_sso_email_lookup

Lookup Sso By Email. Returns: provider_id, organization_id, organization_display_name. Required: email.

list_all_lang_chain_api_keys

List API keys for the current tenant in LangSmith. Returns: created_at, id, short_key, description, read_only, last_used_at, expires_at, workspace_names, default_workspace_name, role_id, org_role_id, access_scope, created_by.

create_a_lang_chain_api_key

Generate a new API key in LangSmith. Returns: created_at, id, short_key, description, read_only, last_used_at, expires_at, workspace_names, default_workspace_name, role_id, org_role_id, access_scope, created_by, key.

delete_a_lang_chain_api_key_by_id

Delete an API key by id in LangSmith. Returns: created_at, id, short_key, description, read_only, last_used_at, expires_at, workspace_names, default_workspace_name, role_id, org_role_id, access_scope, created_by. Required: id.

lang_chain_api_keys_get_current

List personal access tokens for the current user in LangSmith. Deprecated — use /orgs/current/personal-access-tokens instead. Returns: created_at, id, short_key, description, read_only, last_used_at, expires_at, workspace_names, default_workspace_name, role_id, org_role_id, access_scope, created_by.

lang_chain_api_keys_create_current

Generate a personal access token for the current user in LangSmith. Deprecated — use /orgs/current/personal-access-tokens instead. Returns: created_at, id, short_key, description, read_only, last_used_at, expires_at, workspace_names, default_workspace_name, role_id, org_role_id, access_scope, created_by, key, webhooks.

lang_chain_api_keys_delete_current

Delete Personal Access Token. Returns: created_at, id, short_key, description, read_only, last_used_at, expires_at, workspace_names. Required: current_id.

list_all_lang_chain_tenants

List all LangSmith tenants (workspaces) visible to the current authentication. Returns: id, organization_id, created_at, display_name, is_personal, is_deleted, tenant_handle, data_plane_url, read_only, role_id, role_name, permissions, name, handle. Optionally include deleted tenants or skip automatic tenant creation.

create_a_lang_chain_tenant

Create a new LangSmith organization and corresponding workspace (tenant). Returns: id, organization_id, created_at, display_name, is_personal, is_deleted, tenant_handle, data_plane_url, name, handle. Required: display_name.

list_all_lang_chain_onboarding_states

Get the onboarding state for the current user in LangSmith. Returns: id, ls_user_id, tracing_completed_at, lgstudio_completed_at, playground_completed_at, evaluation_completed_at, success_viewed_at, created_at, updated_at.

create_a_lang_chain_onboarding_state

Initialize onboarding state for the current user in LangSmith. Returns: id, ls_user_id, tracing_completed_at, lgstudio_completed_at, playground_completed_at, evaluation_completed_at, success_viewed_at, created_at, updated_at.

lang_chain_onboarding_states_bulk_update

Update a specific onboarding completion field for the current user in LangSmith. Returns: id, ls_user_id, tracing_completed_at, lgstudio_completed_at, playground_completed_at, evaluation_completed_at, success_viewed_at, created_at, updated_at. Required: field.

list_all_lang_chain_ls_user_ids

Get the LangSmith user ID for the current authenticated user. Returns the user ID as a bare string value.

list_all_lang_chain_service_accounts

List service accounts for the current organization in LangSmith. Returns an array of service account objects including id, created_at, and updated_at.

create_a_lang_chain_service_account

Create a service account in LangSmith. Returns the created service account including its id, created_at, and updated_at. Requires a request body. Required: name.

delete_a_lang_chain_service_account_by_id

Delete a LangSmith service account by id. Returns the deleted service account's id, created_at, and updated_at. Required: id.

list_all_lang_chain_workspaces_pendings

List pending workspace invites visible to the current authentication in LangSmith. Returns: id, organization_id, created_at, display_name, is_personal, is_deleted, tenant_handle, data_plane_url.

delete_a_lang_chain_workspaces_pending_by_id

Delete a pending workspace invite in LangSmith by id. Returns an empty 200 response on success. Required: id.

list_all_lang_chain_workspaces_members

List members of the current LangSmith workspace. Returns: tenant_id, members, pending.

create_a_lang_chain_workspaces_member

Add an existing organization member to the current LangSmith workspace. Returns the created member object including identity_id, email, name, role, and created_at.

update_a_lang_chain_workspaces_member_by_id

Update a workspace member in LangSmith, such as changing their assigned role. Returns the updated member object including identity_id, email, name, role, and created_at. Required: id, role_id.

delete_a_lang_chain_workspaces_member_by_id

Remove a member from the current LangSmith workspace by id. Returns an empty 204 response on success. Required: id.

list_all_lang_chain_usage_limits

List the configured usage limits for the current LangSmith tenant. Returns an array of usage limit objects, each containing an id (UUID). Usage limits control the quantity of traces (all traces or extended data retention traces) permitted within a monthly window.

delete_a_lang_chain_usage_limit_by_id

Delete a specific LangSmith usage limit by id. Returns an empty 200 response on success. Required: id.

lang_chain_usage_limits_bulk_update

Upsert a usage limit in LangSmith to control tracing volume (e.g. all traces or extended data retention traces). Returns the usage limit object including its id. A request body specifying the limit configuration is required. Limit values are cached and may take a minute or two before they apply; usage limiting is approximate. Required: limit_type, limit_value.

list_all_lang_chain_usage_limits_orgs

List the configured usage limits for a given organization in LangSmith. Returns an array of UsageLimit objects describing the organization's monthly trace usage limits, including the all-traces limit and the extended data retention traces limit.

list_all_lang_chain_settings

Get the current workspace (tenant) settings in LangSmith. Returns: id, display_name, created_at, tenant_handle, updated_at, thread_id, metadata, status, config.

create_a_lang_chain_settings_handle

Set the tenant handle for the current LangSmith workspace. Returns the updated tenant/workspace object including id and created_at. A JSON request body is required. Required: tenant_handle.

list_all_lang_chain_repos

List all repos in LangSmith with optional filtering by repo type, public visibility, archived status, tags, and more. Returns a paginated list of repos; field details are defined in the ListReposResponse schema in the LangSmith API reference. Max 100 per page.

create_a_lang_chain_repo

Create a new repo in LangSmith. Returns the created repo object; field details are defined in the CreateRepoResponse schema in the LangSmith API reference. Required: repo_handle, is_public.

delete_a_lang_chain_repo_by_id

Delete a repo in LangChain by owner and repo handle. Returns an empty response on success. Required: owner, repo.

lang_chain_repos_bulk_delete

Delete multiple LangSmith repositories by their IDs with partial success support (200 when all deleted, 207 when some fail). Max 100 repo IDs per request. Required: repo_ids.

get_single_lang_chain_repo_by_id

Get a specific LangSmith repo by owner and repo handle. Returns: repo. Required: owner, repo.

lang_chain_repos_list_tags

List all tags for a specific LangSmith repo. Returns: id, repo_id, commit_id, commit_hash, tag_name, created_at, updated_at, name, commit. Required: repo.

lang_chain_repos_create_tag

Create a tag on a LangSmith repo by owner and repo handle. Returns the created tag object including tag_name. Required: owner, repo, tag_name, commit_id.

lang_chain_repos_bulk_update

Update a LangSmith repository identified by owner and repo handle. Returns the updated repository object; field details are defined in the CreateRepoResponse schema in the LangSmith API reference. Required: owner, repo.

create_a_lang_chain_fork

Fork a LangSmith repo to create a copy under your workspace. Returns the forked repo object including id, repo_type, updated_at, is_public, and is_archived. Required: owner, repo, repo_handle.

list_all_lang_chain_repos_tags

List all repo tags in LangSmith, filterable by tenant, repo type, archived status, and visibility. Returns: tags. Max 100 per page.

create_a_lang_chain_repos_optimize_job

Create a prompt optimization job in LangSmith. Returns the optimize prompt response object. Requires a JSON request body conforming to the OptimizePromptJobRequest schema; the field-level breakdown is not enumerated in the available API reference. Required: algorithm, config, prompt_name.

create_a_lang_chain_like

Like a repo in LangSmith by owner and repo name. Returns a like response object whose fields are not individually documented in the available source. Required: owner, repo, like.

lang_chain_likes_like_comment

Like a comment in LangSmith by owner, repo, and parent comment ID. Returns a response object confirming the like action. Required: owner, repo, parent_comment_id.

lang_chain_likes_bulk_delete

Remove a like from a comment in LangSmith. Returns a response object confirming the unlike action. Required: owner, repo, parent_comment_id.

list_all_lang_chain_comments

List comments for a LangSmith tracing project identified by owner and repo. Returns: comments, total. Required: owner, repo.

create_a_lang_chain_comment

Create a comment on a LangSmith tracing project identified by owner and repo. Returns: id. Required: owner, repo, content.

lang_chain_comments_list_replies

List reply comments under a parent comment in a LangSmith tracing project. Returns: comments, total. Required: owner, repo, parent_comment_id.

lang_chain_comments_create_reply

Create a reply comment under a parent comment in a LangSmith tracing project. Returns: id, comment_by, comment_on, parent_id, content, created_at, updated_at, comment_by_name, num_sub_comments, num_likes, liked_by_auth_user. Required: owner, repo, parent_comment_id, content.

list_all_lang_chain_tags

List the transition history for a specific tag in a LangSmith repository, showing each commit change and who performed it. Returns: total, transitions. Required: owner, repo, tag_name.

lang_chain_tags_bulk_update

Update a commit tag in a LangSmith repository, such as moving it to point to a different commit. Returns the updated tag object including id, tag_name, created_at, and updated_at. Required: owner, repo, tag_name, commit_id.

lang_chain_tags_bulk_delete

Delete a tag from a LangSmith repository. Returns an empty 204 response on success. Required: owner, repo, tag_name.

list_all_lang_chain_owners

List all owners of a LangSmith repo. Returns the list of repo owners for the specified repository. Required: owner, repo.

create_a_lang_chain_owner

Add a new owner to a LangSmith repo. Returns the created repo owner object. Required: owner, repo, email.

lang_chain_owners_bulk_delete

Remove an owner from a LangSmith repo. Returns an empty 204 response on success. Required: owner, repo, identity_id.

list_all_lang_chain_optimization_jobs

List all prompt optimization jobs in a LangSmith repository. Returns each job's id, created_at, updated_at. Required: repo.

create_a_lang_chain_optimization_job

Create a new prompt optimization job in a LangSmith repository. Returns: id, repo_id, status, tenant_id, algorithm, config, results, created_at, updated_at. Required: repo, algorithm, config.

update_a_lang_chain_optimization_job_by_id

Replace an existing prompt optimization job with a new, modified job in LangSmith. Returns: id, repo_id, status, tenant_id, algorithm, config, results, created_at, updated_at. Required: id.

delete_a_lang_chain_optimization_job_by_id

Delete a prompt optimization job in LangSmith. Returns an empty 200 response on success. Required: id.

get_single_lang_chain_optimization_job_by_id

Get a specific prompt optimization job by id in LangSmith. Returns: id, repo_id, status, tenant_id, algorithm, config, results, created_at, updated_at, logs. Required: id.

lang_chain_optimization_jobs_create_log

Create a new log entry for a prompt optimization job in LangSmith. Returns the created prompt optimization job log object as defined by the PromptOptimizationJobLog schema. Required: job_id, log_type, message.

list_all_lang_chain_optimization_job_logs

List all logs for a specific prompt optimization job in LangSmith. Returns each log object including its id. Required: job_id.

delete_a_lang_chain_optimization_job_log_by_id

Delete a prompt optimization job log by id in LangSmith. Returns an empty 204 response on success. Required: id.

get_single_lang_chain_optimization_job_log_by_id

Get a specific prompt optimization job log by id in LangSmith. Returns the log object including its id. Required: id.

list_all_lang_chain_well_known_oauth_authorization_servers

Get OAuth2 authorization server metadata in LangSmith per RFC 8414, advertising supported endpoints, grant types, and response types. Returns the RFC 8414 metadata object with spec-defined fields not enumerated in the integration's docs.

list_all_lang_chain_well_known_openid_configurations

Get the OpenID Connect discovery document for LangSmith, advertising the authorization, token, userinfo, and JWKS endpoints plus supported scopes, response types, and signing algorithms. Returns: authorization_endpoint, claims_supported, code_challenge_methods_supported, grant_types_supported, id_token_signing_alg_values_supported, issuer, jwks_uri, response_types_supported, scopes_supported,…

list_all_lang_chain_commits

List commits for a LangSmith repository with pagination. Returns: commits, total. Use '-' as owner for private repos; include_stats controls whether download and view statistics are computed (defaults to true). Required: owner, repo.

create_a_lang_chain_commit

Create a new commit in a LangSmith repository. Returns: commit. Required: owner, repo.

get_single_lang_chain_commit_by_id

Get a specific LangSmith commit by hash, tag, or 'latest'. Returns: commit_hash, description, examples, is_draft, manifest, model_config, model_provider, id, created_at, num_downloads, num_views, model, thread_id, updated_at, metadata, status, config. The commit parameter accepts a hash (8+ chars), a tag name, or 'latest'; hashes shorter than 8 characters are…. Required: owner, repo, commit.

list_all_lang_chain_hub_environments

List the hub environments configuration for the current tenant in LangSmith. Returns: environments, id, webhooks. Returns 404 if no custom configuration exists.

create_a_lang_chain_hub_environment

Create the hub environments configuration for the current tenant in LangSmith. Returns: environments, id, webhooks. Returns 409 Conflict if a configuration already exists. Required: environments.

update_a_lang_chain_hub_environment_by_id

Update a hub environments model in LangSmith by replacing its environments array. Returns: environments, id, webhooks. Required: id, environments.

delete_a_lang_chain_hub_environment_by_id

Delete a hub environments model in LangSmith by id. The tenant reverts to default configuration. Returns a simple key-value response on success. Required: id.

list_all_lang_chain_info

Get information about the current LangSmith deployment in LangChain, including version, instance feature flags, batch-ingest limits, and max SDK versions. Returns: batch_ingest_config, billing_installation_id, customer_info, git_sha, instance_flags, license_expiration_time, sdk_versions, version, batch_ingest_limits, maximum_sdk_versions.

create_a_lang_chain_platform_alert

Create a new alert rule in LangSmith for a given session. Returns the created alert rule including its id and attributes. Required: alert_id, actions, rule.

list_all_lang_chain_platform_alerts

Get a specific alert rule in LangSmith by session and alert rule ID. Returns the alert rule including its id and attributes. Required: session_id, alert_rule_id.

update_a_lang_chain_platform_alert_by_id

Update an alert rule in LangSmith by id. Returns: message. Required: id, session_id, actions, rule.

delete_a_lang_chain_platform_alert_by_id

Delete an alert rule in LangSmith by id. Returns: message. Required: id, session_id.

create_a_lang_chain_alert_test

Test an alert action in LangSmith to verify configuration validity, firing a notification to all configured recipients if valid. Returns: message. Required: session_id.

create_a_lang_chain_item_status

Log the caller's reviewer status for a RUN or THREAD annotation queue item in LangSmith. A null status re-shows the item for this reviewer. Returns: is_archived, override_added_at, queue_item_id, status. Required: queue_item_id.

list_all_lang_chain_annotation_queue_items

List RUN and THREAD items in a LangSmith annotation queue filtered by review status section. Returns: items, next_cursor, previous_cursor. Max page size 100. Required: queue_id, status.

create_a_lang_chain_annotation_queue_item

Add RUN or THREAD items to a LangSmith annotation queue. Returns: items. RUN items require run_id unless created from a suggested example; THREAD items require thread_id and project_id. Required: queue_id.

update_a_lang_chain_annotation_queue_item_by_id

Update mutable timestamps for a LangSmith annotation queue item. Returns: added_at, id, item_type, last_reviewed_time, project_id, queue_id, run_id, source_proposed_example_id, start_time, thread_id. Omit a field or pass null to leave it unchanged. Required: id, queue_id.

list_all_lang_chain_items_counts

Get the annotation queue item count in LangSmith for a specific reviewer or archived bucket. Returns: count. The status parameter must be one of all, needs_my_review, needs_others_review, or archived; start_time and end_time optionally bound the archived item timestamp range. Required: queue_id, status.

create_a_lang_chain_items_delete

Delete RUN or THREAD items from a LangSmith annotation queue by item ID. Returns a map of item identifiers to deletion status strings. Required: queue_id.

list_all_lang_chain_item_placements

Get the placement of a LangSmith annotation queue item, resolving a RUN or THREAD item to its current review section and zero-based position for deep linking. Returns: cursor, item_type, position, section. Required: queue_id, item_id.

create_a_lang_chain_annotation_queue_reviewer

Add a reviewer to a LangSmith annotation queue by assigning a single identity. The operation is idempotent — re-assigning the same identity produces no duplicate. Returns: identity_id. Required: queue_id, identity_id.

delete_a_lang_chain_annotation_queue_reviewer_by_id

Remove a reviewer from a LangChain annotation queue. The operation is idempotent. Returns an empty 204 response on success. Required: id, queue_id.

create_a_lang_chain_examples_delete

Hard delete all versions of dataset example(s) in LangSmith. Inputs, outputs, and metadata are nullified immediately while the example ID, dataset ID, and creation timestamp are retained; attachment files may take up to 7 days to delete. Returns a deletion response. Required: example_ids, hard_delete.

create_a_lang_chain_dataset_example

Upload examples to a LangSmith dataset via multipart/form-data, sending JSON-encoded example data and optional binary attachment files. Returns a 201 Created response confirming the uploaded examples.inputs. Required: dataset_id.

lang_chain_dataset_examples_bulk_update

Update existing examples in a LangSmith dataset via multipart/form-data, sending updated JSON data and attachment operations (retain, rename). Returns a 201 Created response confirming the updated examples. Required: dataset_id.

list_all_lang_chain_platform_evaluators

List evaluators for the current LangSmith workspace, with optional filtering by type, name, tag, feedback key, or resource ID. Returns: evaluators, total. Max 100 per page.

create_a_lang_chain_platform_evaluator

Create a new LLM or code evaluator in LangSmith for the current workspace. Returns the created evaluator including id, name, type, and feedback_key.

lang_chain_platform_evaluators_bulk_delete

Bulk delete multiple evaluators in LangSmith by their IDs, with optional run rule cleanup. Returns per-item success/failure results. Required: evaluator_ids.

get_single_lang_chain_platform_evaluator_by_id

Get a single evaluator in LangSmith by its ID. Returns the evaluator including id, name, type, and feedback_key. Required: id.

update_a_lang_chain_platform_evaluator_by_id

Update an existing evaluator's name, LLM configuration, or code configuration in LangSmith. Returns the updated evaluator including id, name, type, and feedback_key. Returns 409 when a code evaluator build is ENQUEUED or BUILDING. Required: id.

delete_a_lang_chain_platform_evaluator_by_id

Delete a LangChain evaluator by id. Returns an empty 204 response on success. Returns 409 when a code evaluator build is ENQUEUED or BUILDING, or when run rules still reference the evaluator and delete_run_rules is false. Required: id.

list_all_lang_chain_evaluators_spends

Get per-day LLM evaluator spend in LangSmith for a 7-day window starting at period_start. Returns per-day spend data for the requested period, grouped by evaluator, resource, or run rule. Exactly one of group_by, evaluator_id, session_id, or dataset_id must be supplied. Required: period_start.

list_all_lang_chain_platform_features

List feature configurations for the LangSmith workspace. Returns a consolidated view of default models and disabled models per feature. Returns: default_model, disabled_models, feature, org_disabled_providers.

lang_chain_feature_default_models_bulk_update

Set or replace the default model for a feature in a LangChain workspace. Returns an empty 204 response on success. Required: feature.

lang_chain_feature_default_models_bulk_delete

Delete the default model for a feature in the LangChain workspace. Returns an empty 204 response on success. Required: feature.

lang_chain_feature_disabled_models_bulk_update

Disable a model for a feature in LangSmith by adding it to the workspace's disabled list. Returns an empty 204 response on success. Required: feature.

lang_chain_feature_disabled_models_bulk_delete

Re-enable a previously disabled model for a feature in LangSmith by removing it from the workspace's disabled list. Returns an empty 204 response on success. Required: feature, model.

list_all_lang_chain_platform_gateway_policies

List all gateway policies in the current LangSmith organization, including admin-created policies and runtime-materialized children of default policies. Returns each policy with id, name, policy_type, config, subject_matchers, current_spend_usd, and parent_policy_id.

create_a_lang_chain_platform_gateway_policy

Create a gateway policy in LangSmith with a type-specific config. Returns the created policy including id, name, policy_type, config, action, subject_matchers, current_spend_usd, and parent_policy_id. Most types upsert by matchers; route_config and model_fallback require unique names per organization (409 on conflict).

get_single_lang_chain_platform_gateway_policy_by_id

Get a single gateway policy by id in LangSmith. Returns the policy including id, name, policy_type, config, subject_matchers, current_spend_usd, and parent_policy_id. Required: id.

update_a_lang_chain_platform_gateway_policy_by_id

Partially update a gateway policy by id in LangSmith. Only supplied fields are applied; policy_type is immutable. Returns the updated policy including id, name, policy_type, config, subject_matchers, current_spend_usd, and parent_policy_id. Required: id.

delete_a_lang_chain_platform_gateway_policy_by_id

Delete a gateway policy by id in LangSmith. Deleting a default_spend_cap or default_rate_limit also deletes all materialized child policies. Returns an empty 204 response on success. Required: id.

list_all_lang_chain_search_platform_gateway_policies

Search LangSmith gateway policies by a batch of subject matcher values under one key. Returns an array of policy records with id, name, policy_type, config, subject_matchers, action, enabled, priority, current_spend_usd, and parent_policy_id. All request body fields are optional; with subject_matcher_values empty or omitted, returns the same result as the GET list with only policy_type set.

list_all_lang_chain_directories

Get directory contents for a LangChain Hub agent or skill repository, resolving the flattened file tree at a specific commit, tag, or latest. Returns the flattened file tree object with content-specific fields. Use '-' for owner to target the current tenant. Required: owner, repo.

lang_chain_directories_bulk_delete

Delete a LangChain Hub agent or skill repository and its owned child file repositories. Returns an empty 204 response on success. Use '-' for owner to target the current tenant. Required: owner, repo.

create_a_lang_chain_directories_commit

Create directory commit. Returns: commit. Required: owner, repo.

list_all_lang_chain_platform_issues

List LangSmith issues for the authenticated tenant, optionally filtered by session, status, severity, activity, tag, linked trace, or last modified time. Returns: actions, auto_resolution_evidence, auto_resolution_state, created_at, description, first_seen_at, fix_branch, fix_dispatched_at, fix_pr_number, fix_prompt, fix_verification, id, last_seen_at, linear_context, linear_sync, name,…

get_single_lang_chain_platform_issue_by_id

Get a single LangSmith issue by id. Returns: actions, auto_resolution_evidence, auto_resolution_state, created_at, description, first_seen_at, fix_branch, fix_dispatched_at, fix_pr_number, fix_prompt, fix_verification, id, last_seen_at, linear_context, linear_sync, name, proposed_context_fixes, proposed_examples, proposed_fix, proposed_prompt_fixes, recurrences_since_watching, session_id,…. Required: id.

list_all_lang_chain_platform_issues_agents

List all issues agent configs configured for the authenticated tenant in LangSmith. Returns: agent_overview_accepted, analysis_level, auto_open_fix_pr, context_hub_repo_handle, created_at, cron_enabled, cron_schedule, engine_version, github_base_branch, github_repo_subdir, github_repo_url, id, issue_count, latest_run_id, latest_thread_id, linear_available, linear_integration, linear_sync_health,…

create_a_lang_chain_issue_view

Mark a LangSmith issue as viewed by the current user. This beta endpoint records that the user opened the issue; it is idempotent and drives the Engine tab unread-issues badge. Returns an empty 204 response on success. Required: issue_id.

list_all_lang_chain_platform_mcp_vendors

List available MCP vendors in the LangSmith platform catalog. Returns vendor objects including id and attributes with vendor-specific metadata and settings. No required parameters.

get_single_lang_chain_platform_mcp_vendor_by_id

Get a single MCP vendor by id (vendor slug) in LangSmith. Returns the vendor object including id and attributes with vendor-specific metadata and settings. Required: id.

list_all_lang_chain_mcp_vendor_accounts

Get a vendor's account info in LangSmith by resolving the OAuth token for the specified MCP vendor. Returns: organizations, projects. Required: vendor_slug.

list_all_lang_chain_mcp_vendor_mcp_servers

List MCP servers (gateways) for a vendor in LangChain. Returns the MCP gateways available for the workspace's configured org/project. Returns: items, limit, offset, page_count, total_count. Required: vendor_slug.

list_all_lang_chain_mcp_vendor_settings

Get the current MCP vendor settings in LangSmith. Returns the vendor-specific settings object with vendor-specific fields. Required: vendor_slug.

create_a_lang_chain_mcp_vendor_setting

Create (initialize) MCP vendor settings in LangSmith. Returns the vendor-specific settings object with vendor-specific fields. Returns 409 Conflict if settings already exist. Required: vendor_slug.

lang_chain_mcp_vendor_settings_bulk_update

Replace MCP vendor settings in LangSmith. Returns the updated vendor-specific settings object with vendor-specific fields. Required: vendor_slug.

lang_chain_mcp_vendor_settings_bulk_delete

Delete MCP vendor settings in LangSmith. Returns the removed vendor-specific settings object with vendor-specific fields. Required: vendor_slug.

list_all_lang_chain_mcp_vendor_tools

List the tool catalog for a specific MCP vendor in LangChain. Returns: limit, offset, tools, total. Required: vendor_slug.

list_all_lang_chain_oauth_authorized_apps

List authorized OAuth applications in LangSmith that the authenticated user has approved for sign-in. Returns an array of authorized application objects including client_id and additional fields defined in the AuthorizedAppView schema.

delete_a_lang_chain_oauth_authorized_app_by_id

Revoke an authorized OAuth application in LangSmith by id, invalidating that application's active tokens for the user. Returns an empty 204 response on success. Required: id.

list_all_lang_chain_oauth_clients

List OAuth clients owned by the caller's organization in LangSmith. Returns: clients.

create_a_lang_chain_oauth_client

Create a new OAuth 2.0 / OIDC client in LangSmith. For confidential clients, the response includes a client_secret shown only once. Returns: client, client_secret, id.

get_single_lang_chain_oauth_client_by_id

Get a single OAuth client by id in LangSmith. Returns: allowed_scopes, client_id, client_name, client_type, client_uri, created_at, disabled, grant_types, id, logo_uri, policy_uri, redirect_uris, tos_uri, updated_at, client_secret, name, logo, homepage. Required: id.

update_a_lang_chain_oauth_client_by_id

Update an OAuth client in LangSmith by id. Returns the updated client object including id, client_id, created_at, and updated_at. Required: id.

delete_a_lang_chain_oauth_client_by_id

Delete an OAuth client in LangChain by id. Returns an empty 204 response on success. Required: id.

lang_chain_oauth_clients_get_public

Get public OAuth2 client display metadata in LangSmith. Returns the non-sensitive display fields (name, logo, homepage, terms, privacy) used by the consent screen to show a human-readable client identity. Required: client_id.

create_a_lang_chain_client_rotate_secret

Rotate the secret for a LangSmith OAuth client, invalidating the previous secret and generating a new one. Returns: client, client_secret, client_id. The new secret is shown only once. Required: client_id.

create_a_lang_chain_backfills_restart

Restart a backfill job in LangSmith by deleting the backfill job record, causing it to restart from the beginning on the next cron tick. Returns a generic status object with string-valued properties confirming the backfill was restarted. Requires instance admin access.

list_all_lang_chain_orgs_access_policies

List all access policies for the current organization in LangSmith. Returns each access policy record including its id and policy attributes.

create_a_lang_chain_orgs_access_policy

Create a new access policy in LangSmith. Returns the created access policy including its id, created_at, and updated_at. Requires a request body with access policy configuration.

update_a_lang_chain_orgs_access_policy_by_id

Update one or more fields on an access policy in LangSmith. Returns the updated access policy including its id, created_at, and updated_at. Required: id.

delete_a_lang_chain_orgs_access_policy_by_id

Delete a specific access policy by id in LangSmith. Returns an empty 204 response on success. Required: id.

get_single_lang_chain_orgs_access_policy_by_id

Get a specific access policy by id in LangSmith. Returns the access policy including its id, created_at, and updated_at. Required: id.

create_a_lang_chain_role_access_policy

Attach one or more access policies to a role in LangSmith. Returns an empty 204 response on success. Required: role_id.

lang_chain_role_access_policies_attach

Attach one or more access policies to a role in LangSmith. Returns an empty 204 response on success. Required: role_id.

lang_chain_role_access_policies_bulk_delete

Detach one or more access policies from a role in LangSmith. Returns an empty 204 response on success. Required: role_id, access_policy_ids.

list_all_lang_chain_scim_tokens

List all SCIM bearer tokens for the current organization in LangSmith. Returns: created_at, created_by, description, id, last_used_at, short_token, updated_at. Full token values are not returned.

create_a_lang_chain_scim_token

Create a new SCIM bearer token for the current organization in LangSmith. Returns: created_at, created_by, description, id, last_used_at, short_token, token, updated_at. The full token value is only returned once upon creation.

update_a_lang_chain_scim_token_by_id

Update the description of an existing SCIM token in LangSmith. Returns: created_at, created_by, description, id, last_used_at, short_token, updated_at. Required: id.

delete_a_lang_chain_scim_token_by_id

Delete a SCIM bearer token by id in LangSmith. Returns an empty 204 response on success. Required: id.

get_single_lang_chain_scim_token_by_id

Get a single SCIM bearer token by id in LangSmith. Returns: created_at, created_by, description, id, last_used_at, short_token, updated_at. The full token value is not returned. Required: id.

list_all_lang_chain_session_agent_versions

List all agent versions (commit SHAs) observed in a LangSmith tracing project, ordered by first_seen_at descending. Returns: commit_sha, first_seen_at, id. Required: session_id.

list_all_lang_chain_session_issues_agents

Get the issues agent config attached to a LangSmith tracer session. Returns the config object including id, agent_overview_repo_id, and additional configuration attributes. Beta endpoint — may change without notice. Required: session_id.

create_a_lang_chain_session_issues_agent

Create and configure the issues agent for a LangSmith tracer session, enqueueing the initial scan. Returns the created issues agent config object including id, agent_overview_repo_id, and additional configuration attributes. Fails with 409 if an agent already exists. Beta endpoint — may change without notice. Required: session_id.

lang_chain_session_issues_agents_bulk_update

Update the issues agent config for a LangSmith tracer session. Patches the config; omitted fields are left unchanged and side effects (setting agent_overview_repo_id) run in a single transaction. Returns the updated agent config including id, agent_overview_repo_id, and additional configuration attributes. Beta endpoint — may change without notice. Required: session_id.

lang_chain_session_issues_agents_bulk_delete

Delete the issues agent for a LangSmith tracer session, removing the agent config, its issues, and the agent-overview hub repo. Returns an empty 204 response on success. Beta endpoint — may change without notice. Required: session_id.

lang_chain_issues_agent_overviews_bulk_update

Save the issues agent overview for a tracer session in LangSmith, creating or updating the backing private Prompt Hub repo and linking it to the issues agent config. Returns a JSON response whose exact fields are defined by the upstream agent.SaveOverviewResponse schema (not fully documented in the discovered source). Beta endpoint — may change without notice. Required: session_id.

create_a_lang_chain_webhook_roll_secret

Roll the signing secret for a LangSmith issues agent webhook, replacing it with a new one that takes effect immediately for future deliveries. Returns: created_at, destination_type, event_types, has_jira_token, has_signing_secret, has_unreadable_credentials, header_names, headers, id, issue_statuses, organization_id, session_id, severity_threshold, signing_secret, slack_channel_id, slack_team_id,…. Required: session_id, webhook_id.

list_all_lang_chain_issues_views

List viewed issues for a LangSmith session — issues the current user has opened, with view timestamps. Returns: views. Beta endpoint; may change without notice. Required: session_id.

list_all_lang_chain_platform_tools

List tools in the LangSmith workspace with optional search filtering. Returns each tool's id, handle, name, and description. The optional query parameter filters tools by name or description.

create_a_lang_chain_platform_tool

Create a new tool in the LangSmith workspace. Returns the created tool including its id, handle, name, and description. Required: description, handle, name, parameters.

get_single_lang_chain_platform_tool_by_id

Get a single tool by its handle in LangSmith. Returns: created_at, description, enabled, handle, id, metadata, name, returns, tenant_id, updated_at, webhooks. Required: handle.

lang_chain_platform_tools_bulk_update

Update an existing tool by its handle in LangSmith. Returns the updated tool including its id, handle, name, and description. Required: handle.

lang_chain_platform_tools_bulk_delete

Delete a tool by its handle in LangSmith. Returns an empty 204 response on success. Required: handle.

get_single_lang_chain_tools_id_by_id

Get a single LangSmith tool by its UUID. Returns: created_at, description, enabled, handle, id, metadata, name, returns, tenant_id, updated_at. Required: id.

update_a_lang_chain_tools_id_by_id

Update an existing LangSmith tool by its UUID. Returns the updated tool including id, name, description, and handle. Required: id.

delete_a_lang_chain_tools_id_by_id

Delete a LangChain tool by its UUID. Returns an empty 204 response on success. Required: id.

create_a_lang_chain_experiment_run

Fetch experiment runs for a publicly shared LangSmith dataset by share token. Returns a paginated page of dataset examples with their experiment runs. Self-hosted deployments require LangSmith v0.16 or later. Required: share_token.

create_a_lang_chain_dataset_experiment_run

Fetch experiment runs for dataset examples in LangSmith by posting experiment selection criteria. Returns a paginated page of examples with their associated runs from the requested experiments. Returns: items, next_cursor. Self-hosted deployments require LangSmith v0.16 or later. Required: dataset_id.

get_single_lang_chain_public_thread_by_id

Get a public shared thread manifest in LangSmith by share token. Returns the thread and project the share token is scoped to, plus the project's conversations config, including thread_id, created_at, updated_at, status, metadata, config, and values. Required: id.

list_all_lang_chain_public_threads

Stream messages for a public shared LangSmith thread as server-sent events (SSE). Returns an SSE stream (text/event-stream) containing the thread's conversation messages. Required: share_token.

list_all_lang_chain_thread_runs

Get one run in a public shared thread in LangSmith. Returns a single run including its id, name, run_type, status, start_time, end_time, inputs, and outputs. Required: share_token, run_id, start_time, selects.

list_all_lang_chain_thread_stats

Get aggregate stats for a publicly shared thread in LangSmith by its share token. Returns: completion_cost, completion_cost_details, completion_token_details, completion_tokens, feedback_stats, first_start_time, last_end_time, last_start_time, latency_p50_seconds, latency_p99_seconds, prompt_cost, prompt_cost_details, prompt_token_details, prompt_tokens, total_cost, total_tokens, turns,…. Required: share_token, selects.

get_single_lang_chain_thread_stat_by_id

Get aggregate stats for a single thread in LangSmith within a project, including turn count, latency percentiles, token/cost sums, and detail breakdowns. Returns: completion_cost, completion_cost_details, completion_token_details, completion_tokens, feedback_stats, first_start_time, last_end_time, last_start_time, latency_p50_seconds, latency_p99_seconds, prompt_cost, prompt_cost_details,…. Required: thread_id, selects, session_id.

list_all_lang_chain_thread_traces

List root traces in a public shared LangSmith thread by share token. Returns: items, next_cursor. Each trace in items carries selectable fields such as trace_id, name, start_time, end_time, latency, total_tokens, and total_cost. Max 100 per page. Required: share_token, selects.

lang_chain_thread_traces_list_private

List all traces belonging to a specific thread within a LangSmith project. Returns: items, next_cursor, thread_id, created_at, updated_at, metadata, status, config. Each trace in items carries selectable fields such as trace_id, name, start_time, end_time, latency, total_tokens, and total_cost. Max 100 per page. Required: thread_id, project_id.

list_all_lang_chain_trace_runs

List runs of one trace in a publicly shared LangSmith thread. Returns run objects containing only the fields requested via the selects parameter — e.g., id, name, run_type, status, start_time, end_time, inputs, outputs. Self-hosted deployments require LangSmith v0.16 or later. Required: share_token, trace_id, selects.

lang_chain_trace_runs_list_private

List runs in a LangSmith trace, optionally filtered by time range and filter expression. Returns run objects with the fields requested via selects (defaults to id only) — e.g., id, name, run_type, status, start_time, end_time, inputs, outputs. Required: trace_id, project_id.

list_all_lang_chain_run_urls

Get the LangSmith UI URL for a specific run in LangChain. Returns: url. Self-hosted deployments require LangSmith v0.16 or later. Required: run_id, project_id, trace_id.

list_all_lang_chain_sandboxes_boxes

List LangSmith sandboxes with optional filtering by name substring, status, creator, and labels, plus sorting. Returns: items, next_cursor, offset, sandboxes.

create_a_lang_chain_sandboxes_box

Create a new LangSmith sandbox from a snapshot. Provide at most one of snapshot_id or snapshot_name; if neither is given the default snapshot is used. Returns: cpu_millicores, created_at, created_by, dataplane_url, delete_after_stop_seconds, fs_capacity_bytes, id, idle_ttl_seconds, labels, mem_bytes, mount_config, name, preserve_memory_on_stop, proxy_config, size_class, snapshot_id, status,…

lang_chain_sandboxes_boxes_start

Start a stopped or failed LangSmith sandbox. This endpoint is not idempotent. Returns: cpu_millicores, created_at, created_by, dataplane_url, delete_after_stop_seconds, fs_capacity_bytes, id, idle_ttl_seconds, labels, mem_bytes, mount_config, name, preserve_memory_on_stop, proxy_config, size_class, snapshot_id, status, status_message, stopped_at, updated_at, updated_by, vcpus. Required: name.

lang_chain_sandboxes_boxes_stop

Stop a ready sandbox in LangChain by name. The filesystem is preserved for later restart and the endpoint is not idempotent. Returns an empty 204 response on success. Required: name.

get_single_lang_chain_sandboxes_box_by_id

Get a LangSmith sandbox by display name. Stale provisioning sandboxes are auto-failed. Returns: cpu_millicores, created_at, created_by, dataplane_url, delete_after_stop_seconds, fs_capacity_bytes, id, idle_ttl_seconds, labels, mem_bytes, mount_config, name, preserve_memory_on_stop, proxy_config, size_class, snapshot_id, status, status_message, stopped_at, updated_at, updated_by, vcpus,…. Required: name.

lang_chain_sandboxes_boxes_bulk_update

Update a LangSmith sandbox's display name, retention, resources, tags, or proxy configuration. Proxy config sent to a non-running sandbox is stored and applied on next start. Returns: cpu_millicores, created_at, created_by, dataplane_url, delete_after_stop_seconds, fs_capacity_bytes, id, idle_ttl_seconds, labels, mem_bytes, mount_config, name, preserve_memory_on_stop, proxy_config, size_class,…. Required: name.

lang_chain_sandboxes_boxes_bulk_delete

Delete a LangSmith sandbox by name or UUID, tearing down the sandbox runtime and removing the DB record. Returns an empty 204 response on success. Required: name.

create_a_lang_chain_boxes_batch_delete

Batch delete multiple LangChain sandboxes by name or UUID in a single request. Returns a confirmation response indicating the outcome of the deletion.

create_a_lang_chain_boxe_download_url

Generate a tokenized download link for a single file in a LangSmith sandbox. The link requires no further authentication and pins the sandbox and file path so it cannot be repointed at another file. Returns the download URL response object with fields defined in the upstream sandboxes.DownloadURLResponse schema. Links never expire unless expires_in_seconds is set in the…. Required: name, path.

create_a_lang_chain_boxe_service_url

Generate a short-lived JWT service access token for an HTTP service running inside a LangSmith sandbox. Returns: browser_url, expires_at, service_url, token, expiry. Required: name.

create_a_lang_chain_boxe_snapshot

Capture a snapshot from a LangChain sandbox by capturing its current state or promoting an existing checkpoint. Returns: created_at, created_by, description, docker_image, fs_capacity_bytes, fs_used_bytes, id, image_digest, labels, memory_snapshot_size_bytes, name, registry_id, source_sandbox_id, status, status_message, tags, updated_at. Required: name.

list_all_lang_chain_boxe_status

Get the lightweight status of a LangChain sandbox for polling. Returns a sandbox status object with its id and status attributes. Required: name.

list_all_lang_chain_sandboxes_registries

List LangSmith sandbox registries for pulling private images. Returns a list of registry objects with fields defined by the RegistryListResponse component schema; consult the upstream OpenAPI spec for the field-level breakdown. Optional filter: name_contains.

create_a_lang_chain_sandboxes_registry

Create a LangSmith sandbox registry for pulling private images. Returns the created registry object with fields defined by the RegistryResponse component schema. A JSON request body defined by the CreateRegistryPayload schema is required; consult the upstream OpenAPI spec for the field-level breakdown. Required: name, password, url, username.

get_single_lang_chain_sandboxes_registry_by_id

Get a single LangSmith sandbox registry by name. Returns the registry object with fields defined by the RegistryResponse component schema; consult the upstream OpenAPI spec for the field-level breakdown. Required: name.

lang_chain_sandboxes_registries_bulk_update

Update a LangSmith sandbox registry's name and/or credentials. Returns the updated registry object with fields defined by the RegistryResponse component schema; consult the upstream OpenAPI spec for the field-level breakdown. Required: name.

lang_chain_sandboxes_registries_bulk_delete

Delete a LangSmith sandbox registry by name. Returns an empty 204 response on success. Required: name.

list_all_lang_chain_sandboxes_snapshots

List LangChain sandbox snapshots with optional filtering by name, status, creator, and labels. Returns: items, next_cursor, offset, snapshots. Supports sorting by name, status, or created_at.

create_a_lang_chain_sandboxes_snapshot

Create a LangChain sandbox snapshot from a Docker image (async build). Returns: created_at, created_by, description, docker_image, fs_capacity_bytes, fs_used_bytes, id, image_digest, labels, memory_snapshot_size_bytes, name, registry_id, source_sandbox_id, status, status_message, tags, updated_at. Required: docker_image, fs_capacity_bytes, name.

get_single_lang_chain_sandboxes_snapshot_by_id

Get a LangSmith sandbox snapshot by id, accepting a snapshot UUID, a name:tag reference, or a bare snapshot name (which resolves to name:latest). Returns: created_at, created_by, description, docker_image, fs_capacity_bytes, fs_used_bytes, id, image_digest, labels, memory_snapshot_size_bytes, name, registry_id, source_sandbox_id, status, status_message, tags, updated_at. Required: id.

delete_a_lang_chain_sandboxes_snapshot_by_id

Delete a LangSmith sandbox snapshot by id, accepting a snapshot UUID or a Docker-style name[:tag] reference. The underlying storage is reclaimed asynchronously. Returns an empty 204 response on success. Required: id.

list_all_lang_chain_sandboxes_snapshots_by_names

Get a LangSmith sandbox snapshot by name, returning every tag under that name and the snapshot each tag resolves to. Returns: name, tags. Required: name.

list_all_lang_chain_sandboxes_usages

Get current sandbox resource usage and quota limits for the LangSmith workspace. Returns a usage object whose fields are defined by the LangSmith sandboxes UsageResponse schema; consult the API reference for the field-level breakdown. No required parameters.

list_all_lang_chain_sandboxe_downloads

Download file contents from a LangChain sandbox filesystem path. Returns the binary file content as an octet-stream; supports HTTP range requests for partial downloads. Required: sandbox_id, path.

lang_chain_sandboxes_execute

Execute a command inside a LangSmith sandbox. Returns: exit_code, stderr, stdout. Use the streaming execute endpoints for long-running commands that may exceed the synchronous request deadline. Required: sandbox_id.

create_a_lang_chain_stream_resume

Resume streaming a sandbox command in LangSmith that was started by the stream start endpoint. The request body offsets acknowledge output already received, freeing the sandbox buffer and unpausing a waiting command. Attaches only to an existing command; a command the sandbox no longer has returns 404. Returns an SSE event stream of stdout/stderr output with base64 payloads. Required: sandbox_id.

create_a_lang_chain_stream_start

Start executing a command in a LangChain sandbox, streaming stdout/stderr as Server-Sent Events with base64 payloads. Returns an SSE event stream whose events include stdout, stderr, and ack_required; the stream ends with an ack_required event when the sandbox output buffer needs an ack. Requires a v2 runtime sandbox; pass command_id in the body to reuse a running command…. Required: sandbox_id.

list_all_lang_chain_execute_ws

Open a WebSocket connection for streaming command execution inside a LangChain sandbox. Returns a WebSocket upgrade (101 Switching Protocols) with no HTTP response body. Required: sandbox_id.

create_a_lang_chain_sandboxe_glob

Glob a LangSmith sandbox filesystem to find files under a root path matching a glob pattern (supports **). Returns matching file entries in lexical order by path. Required: sandbox_id.

create_a_lang_chain_sandboxe_grep

Search files under a root path in a LangChain sandbox for a literal text pattern (not a regex). Returns the grep response from the sandbox filesystem; field names are defined in the upstream sandboxes.GrepResponse schema and were not resolved in the discovered source. Required: sandbox_id.

list_all_lang_chain_sandboxe_tunnels

Open a WebSocket tunnel to a specific TCP port inside a LangChain sandbox. On success, returns a 101 WebSocket upgrade with no JSON response body. Required: sandbox_id.

create_a_lang_chain_sandboxe_upload

Upload a file to a LangChain sandbox filesystem at a specified destination path. Returns the upload response object (attributes) defined by the upstream UploadResponse schema. Required: sandbox_id, path.

create_a_lang_chain_threads_query

Query threads within a LangSmith project (session) using cursor-based pagination with time range and optional filters. Returns: items, next_cursor. Self-hosted deployments require LangSmith v0.16 or later.

list_all_lang_chain_thread_shares

Read a thread's share state in LangSmith, returning the share token or 404 when the thread is not shared. Returns: share_token. Required: thread_id, project_id.

create_a_lang_chain_thread_share

Share a thread in LangSmith by minting a public share token. Idempotent — sharing an already-shared thread returns the existing token. Returns: share_token. Required: thread_id, project_id.

lang_chain_thread_shares_bulk_delete

Delete the share token for a LangChain thread, unsharing it. The operation is idempotent — it Returns an empty 204 response on success. Returns an empty 204 response on success. Required: thread_id, project_id.

create_a_lang_chain_traces_query

Query traces (root runs) for a single LangSmith tracing project within a start-time window. Returns a paginated list of traces, each carrying the root run plus trace_aggregates (total_tokens, total_cost, first_token_time). Supports trace_filter, tree_filter, selects, and cursor pagination. min_start_time defaults to 24 hours before the request and max_start_time defaults to…

list_all_lang_chain_auth_publics

Get public authentication information for the current LangSmith workspace-level session. Returns the authn.PublicAuthInfo object whose field set is defined upstream and not enumerated in the available documentation.

create_a_lang_chain_aws_marketplace_register

Register an AWS Marketplace customer by submitting the registration token received when a customer clicks 'Set Up Account' in AWS Marketplace. Resolves the customer identity, stores it, and issues a redirect. Returns a 303 redirect to the thank-you page with no response body.

list_all_lang_chain_dataset_experiment_view_overrides

List experiment view override configurations for a LangSmith dataset. Returns: column_overrides, created_at, dataset_id, id, modified_at. Each override entry includes a column name, color_gradient, precision, and hide flag. Required: dataset_id.

create_a_lang_chain_dataset_experiment_view_override

Create a new experiment view override configuration for a LangSmith dataset. Returns: column_overrides, created_at, dataset_id, id, modified_at. Each override entry includes a column name, color_gradient, precision, and hide flag. Fails if an override already exists for the dataset; use update to modify an existing one. Required: dataset_id, column_overrides.

get_single_lang_chain_dataset_experiment_view_override_by_id

Get a single experiment view override configuration by id in LangSmith. Returns: column_overrides, created_at, dataset_id, id, modified_at. Each override entry includes a column name, color_gradient, precision, and hide flag. Required: id, dataset_id.

update_a_lang_chain_dataset_experiment_view_override_by_id

Update an existing experiment view override configuration in LangSmith by completely replacing its column overrides. Returns: column_overrides, created_at, dataset_id, id, modified_at. Each override entry includes a column name, color_gradient, precision, and hide flag. Required: id, dataset_id, column_overrides.

delete_a_lang_chain_dataset_experiment_view_override_by_id

Delete an experiment view override configuration by id in LangSmith. Returns an empty 204 response on success. Required: id, dataset_id.

list_all_lang_chain_providers

Get the authenticated user's provider user ID for a given provider type in LangSmith. Returns the provider user ID as a string, or null if not set. Required: provider_type.

list_all_lang_chain_oauth_authorizes

Initiate OAuth2 authorization in LangChain by validating request parameters and redirecting to the consent page. Returns a 302 redirect with no response body on success. Required: response_type, client_id, redirect_uri, code_challenge, code_challenge_method.

create_a_lang_chain_authorize_approve

Approve an OAuth2 authorization request in LangSmith, issuing an authorization code after the user consents. Returns: redirect_uri.

create_a_lang_chain_device_authorize

Authorize a device code in LangChain, marking it as authorized for the authenticated user during the OAuth2 device flow. Returns a JSON object with string-valued properties confirming the authorization.

create_a_lang_chain_device_code

Request OAuth2 device authorization in LangSmith per RFC 8628, issuing a device code and user code for the device authorization flow. Returns: device_code, expires_in, interval, user_code, verification_uri.

create_a_lang_chain_oauth_register

Register an OAuth2 dynamic client in LangSmith via RFC 7591 Dynamic Client Registration. Returns the created client registration response object. Only mints public clients with allowed loopback, HTTPS, or native client redirect URIs; body limit 8 KB.

create_a_lang_chain_oauth_revoke

Revoke an OAuth2 access or refresh token in LangChain per RFC 7009. Returns an empty 200 response on success regardless of whether the token was found.

create_a_lang_chain_oauth_token

Exchange a grant for OAuth2 tokens in LangSmith. The endpoint dispatches by grant_type: authorization_code (with PKCE code_verifier), urn:ietf:params:oauth:grant-type:device_code, or refresh_token. Returns an OAuth2 token response whose fields are not individually documented in the available source.

list_all_lang_chain_orgs_data_planes

List data planes for the current organization in LangSmith. Returns up to 50 data planes sorted by status priority (active first), then newest first. Returns: data_planes. Requires a BYOC-enabled organization.

create_a_lang_chain_orgs_data_plane

Create a new data plane in LangSmith. Persists the rendered data plane spec and returns the data plane with status=requested. Returns: api_url, created_at, firewall, id, maintenance_window, name, region, status, status_updated_at, ttl, workspaces, updated_at. Requires a BYOC-enabled organization with org admin permissions.

update_a_lang_chain_orgs_data_plane_by_id

Update specific settings for a data plane in LangSmith owned by the caller's organization. Returns: api_url, created_at, firewall, id, maintenance_window, name, region, status, status_updated_at, ttl, workspaces, updated_at. Required: id.

delete_a_lang_chain_orgs_data_plane_by_id

Delete a data plane in LangSmith by id. Verifies AWS role delete permissions, removes linked workspaces, and starts asynchronous deprovisioning. Returns: api_url, created_at, firewall, id, maintenance_window, name, region, status, status_updated_at, ttl, workspaces, updated_at. Requires a BYOC-enabled organization with org admin permissions. Required: id.

list_all_lang_chain_agent_builder_integrations

Get Agent Builder integrations settings for the current workspace in LangSmith. Returns the integrations settings payload including default policy, integration overrides, and known integrations.

lang_chain_agent_builder_integrations_bulk_update

Replace Agent Builder integrations settings for the current workspace in LangSmith. Returns the updated integrations settings payload including default policy, integration overrides, and known integrations.

list_all_lang_chain_fleet_orgs

List LangSmith Fleet organizations the authenticated caller belongs to. Returns: items, next_cursor, orgs. This endpoint is the Fleet bootstrap entry point — use an organization's id to list its workspaces via the tenants endpoint. Max 20 per page.

list_all_lang_chain_org_tenants

List LangSmith tenants (workspaces) visible to the authenticated caller in a given organization, intended for Fleet bootstrap. Returns each tenant's id. Max 20 items per page. Required: org_id.

list_all_lang_chain_fleet_secrets

List workspace secret names in LangChain to check which model API keys are already configured. Returns: items, next_cursor. Secret values are never returned. Max 100 per page.

create_a_lang_chain_fleet_secret

Bulk set or delete workspace secrets in LangChain in a single request. A null value deletes the key; a non-null value sets it. Returns an empty 204 response on success.

lang_chain_fleet_secrets_bulk_update

Create or update a single workspace secret by name in LangChain. The value is write-only and never returned by any endpoint. Returns an empty 204 response on success. Required: name.

lang_chain_fleet_secrets_bulk_delete

Delete a workspace secret by name in LangChain. Succeeds whether or not the secret currently exists. Returns an empty 204 response on success. Required: name.

get_single_lang_chain_tenant_user_by_id

Get a LangSmith fleet user by ID within a specific tenant workspace. Returns: id, name. Required: id, tenant_id.

create_a_lang_chain_thread_sandbox_activation

Activate a thread sandbox in LangSmith. Starts or resumes the sandbox referenced by the thread and returns when it is ready. Returns: sandbox_slug, scope, status, sandbox_id. The operation is idempotent; the thread must include sandbox.sandbox_slug. Required: thread_id.

list_all_lang_chain_fleet_users

List Fleet users (workspace members) in LangSmith, optionally filtered by a search string matched against member email and full name. Returns: items, next_cursor. Max 100 per page.

get_single_lang_chain_fleet_user_by_id

Get the current authenticated Fleet user's profile in LangSmith. Returns: email, id, name.

list_all_lang_chain_workspaces_ttl_settings

Get the longlived trace TTL settings for the current workspace in LangSmith. Returns: is_custom, longlived_ttl_days, tenant_id.

lang_chain_workspaces_ttl_settings_bulk_update

Update the longlived trace TTL for the current workspace in LangSmith. Returns: is_custom, longlived_ttl_days, tenant_id.

Why Truto

Why use Truto’s MCP server for LangChain

Other MCP servers give you a static tool list for one app. Truto gives you a managed, multi-tenant MCP infrastructure across 800+ integrations.

01

Auto-generated, always up to date

Tools are dynamically generated from curated documentation — not hand-coded. As integrations evolve, tools stay current without manual maintenance.

02

Fine-grained access control

Scope each MCP server to read-only, write-only, specific methods, or tagged tool groups. Expose only what your AI agent needs — nothing more.

03

Multi-tenant by design

Each MCP server is scoped to a single connected account with its own credentials. The URL itself is the auth token — no shared secrets, no credential leaking across tenants.

04

Works with every MCP client

Standard JSON-RPC 2.0 protocol. Paste the URL into Claude, ChatGPT, Cursor, or any MCP-compatible agent framework — tools are discovered automatically.

05

Built-in auth, rate limits, and error handling

Tool calls execute through Truto’s proxy layer with automatic OAuth refresh, rate-limit handling, and normalized error responses. No raw API plumbing in your agent.

06

Expiring and auditable servers

Create time-limited MCP servers for contractors or automated workflows. Optional dual-auth requires both the URL and a Truto API token for high-security environments.

How It Works

From zero to integrated

Go live with LangChain in under an hour. No boilerplate, no maintenance burden.

01

Link your customer’s LangChain account

Use Truto’s frontend SDK to connect your customer’s LangChain account. We handle all OAuth and API key flows — you don’t need to create the OAuth app.

02

We handle authentication

Don’t spend time refreshing access tokens or figuring out secure storage. We handle it and inject credentials into every API request.

03

Call our API, we call LangChain

Truto’s Proxy API is a 1-to-1 mapping of the LangChain API. You call us, we call LangChain, and pass the response back in the same cycle.

04

Unified response format

Every response follows a single format across all integrations. We translate LangChain’s pagination into unified cursor-based pagination. Data is always in the result attribute.

FAQs

Common questions about LangChain on Truto

Authentication, rate limits, data freshness, and everything else you need to know before you integrate.

What LangChain product does this integration actually talk to?

The integration targets the LangSmith API — the commercial LLMOps platform in the LangChain ecosystem. It covers runs, traces, threads, feedback, datasets, examples, annotation queues, workspaces, evaluators, and billing/usage data.

How does authentication work for end users?

End users connect by providing a LangSmith API key or service key scoped to their workspace or organization. Truto stores and rotates the credential per connected account, so your app never handles raw secrets in code.

Can I write feedback back to specific LangSmith runs?

Yes. You can create, update, and delete feedback tied to a specific run ID, including scores, comments, and correction values. Bulk and public feedback endpoints are also supported for higher-throughput use cases.

Does the integration support multi-tenant workspaces?

Yes. You can list, create, and manage LangSmith organizations and workspaces, provision members, set per-workspace TTL and usage limits, and issue scoped service keys — which is essential when your customers separate environments or client tenants.

What data is available for cost and usage reporting?

You can list billing usage, granular usage, and workspace usage limits, plus thread stats and run-level token/latency metrics. This is enough to build cost dashboards, chargeback reports, and automated spend guardrails.

Can I manage datasets and evaluations programmatically?

Yes. The integration exposes datasets, dataset versions, examples (including bulk create/update/delete and uploads), dataset splits, comparative experiments, and platform evaluators, so you can build full evaluation workflows and CI checks on top.

Are annotation queues and human-in-the-loop workflows supported?

Yes. You can create annotation queues, populate them with runs, list and update queue items, add reviewers, and export results — enough to build a custom labeling UI backed by LangSmith.

LangChain

Get LangChain integrated into your app

Our team understands what it takes to make a LangChain integration successful. A short, crisp 30 minute call with folks who understand the problem.