Connect Sentry to ChatGPT: Monitor Errors and App Performance

If you need to connect Sentry to ChatGPT to automate error triage, monitor application performance, or analyze stack traces, you need a Model Context Protocol (MCP) server. This server acts as the translation layer between ChatGPT's tool calls and Sentry's complex REST APIs. You can either build and maintain this JSON-RPC infrastructure yourself, or use a managed integration platform like Truto to dynamically generate a secure, authenticated MCP server URL.

If your team uses Claude, check out our guide on connecting Sentry to Claude or explore our broader architectural overview on connecting Sentry to AI Agents.

Giving a Large Language Model (LLM) read and write access to an application monitoring platform like Sentry is an engineering challenge. You have to handle complex hierarchical data models (projects, issues, events, stack traces), map these nested structures to flat MCP tool definitions, and deal with strict pagination constraints. Every time your engineering team wants to expose a new metric or event type, custom server code must be updated, redeployed, and tested.

This guide breaks down exactly how to use Truto to generate a secure, managed MCP server for Sentry, connect it natively to ChatGPT, and execute complex incident response workflows using natural language.

The Engineering Reality of the Sentry API

A custom MCP server is essentially a self-hosted API gateway. While the open MCP standard provides a predictable way for models to discover tools, implementing it against Sentry's API requires significant domain knowledge.

If you decide to build a custom MCP server for Sentry, you own the entire lifecycle of the integration. Here are the specific integration challenges that break standard CRUD assumptions when working with Sentry:

The Issue vs. Event Hierarchy

Sentry does not treat "errors" as a flat list. An error occurrence is an Event. Sentry groups similar events together into an Issue (internally called a Group) using a complex fingerprinting and hashing algorithm. When an LLM asks "What is causing the database timeout error?", it cannot just query a list of errors. The MCP server must first query the Issues endpoint, find the relevant Issue ID, and then query the Events bound to that specific Issue to retrieve the actual stack trace.

Pagination Constraints and Payload Inflation

When pulling event details, Sentry offers a full=true query parameter to include complete stack traces and environmental metadata. However, passing this parameter strictly caps the pagination size to a maximum of 10 results per page to prevent payload bloat. If your custom MCP server requests 50 full events, the Sentry API will reject or truncate the request. Your MCP tool definitions must explicitly teach the LLM about these parameter constraints, otherwise ChatGPT will hallucinate query arguments that Sentry rejects.

Dual Scoping: Organizations vs. Projects

Sentry's API endpoints are heavily bifurcated. Some resources (like Teams, Members, and top-level Stats) are scoped to the organization_slug. Other resources (like Releases, Envrionments, and specific Events) are scoped to both the organization_slug and the project_slug. Your MCP tool schemas must force the LLM to provide the correct routing identifiers for every single call, which often requires chained tool executions just to resolve a project name to its internal slug.

Sentry to ChatGPT Quickstart Guide

If you just want the fastest path from a fresh Truto account to ChatGPT calling the Sentry API, follow these steps.

What you need:

• A Truto account with API access.
• A Sentry organization admin account for OAuth consent.
• A ChatGPT Pro, Plus, Business, Enterprise, or Education seat with Developer mode available.

Step 1: Connect Sentry as an Integrated Account

First, you need to establish the connection between Truto and Sentry. In the Truto dashboard, open Integrated Accounts -> New Integrated Account, select Sentry, and run the OAuth flow. Truto securely stores the refresh token and automatically refreshes access tokens before they expire, ensuring ChatGPT never encounters an expired credential.

Step 2: Generate the Sentry MCP Server

Truto scopes each MCP server to a single integrated account. You can create this server via the Truto UI or programmatically via the API.

Method A: Via the Truto UI

1. Navigate to the integrated account page for your Sentry connection.
2. Click the MCP Servers tab.
3. Click Create MCP Server.
4. Select your desired configuration (e.g., read-only methods, specific tags).
5. Copy the generated MCP server URL.

Method B: Via the API You can generate the server dynamically using a single POST request. This is ideal for multi-tenant applications provisioning agent access on the fly.

curl -X POST https://api.truto.one/integrated-account/$INTEGRATED_ACCOUNT_ID/mcp \
  -H "Authorization: Bearer $TRUTO_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Sentry DevOps Agent",
    "config": {
      "methods": ["read", "write"]
    }
  }'

The response contains a url field structured as https://api.truto.one/mcp/<token>. This URL handles all routing, authentication, and token refreshing. Treat this URL as a secret.

Step 3: Connect the MCP Server to ChatGPT

Now, you must register this endpoint with ChatGPT.

Method A: Via the ChatGPT UI

1. In ChatGPT, navigate to Settings -> Apps -> Advanced settings.
2. Enable Developer mode.
3. Under MCP servers / Custom connectors, click to add a new server.
4. Set the Name to "Sentry (Truto)".
5. Paste the Truto MCP URL into the Server URL field and click Add.

Method B: Via Manual Config File If you are running a local agent orchestration framework or a headless client compatible with MCP, you can define the server connection using Server-Sent Events (SSE) via the standard MCP CLI transport.

{
  "mcpServers": {
    "sentry_prod": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-sse",
        "--url",
        "https://api.truto.one/mcp/a1b2c3d4e5f6..."
      ]
    }
  }
}

Once connected, ChatGPT will automatically request the tools/list endpoint, and Truto will dynamically generate the JSON-RPC schemas based on Sentry's API documentation.

Hero Tools for Sentry

Truto exposes the full surface area of the Sentry API to your LLM, but these specific tools represent the highest-leverage operations for AI-driven incident response and observability.

List All Organization Issues

Tool: list_all_sentry_organization_issues

This tool retrieves a list of active issues (error groups) for a given organization. It serves as the primary entry point for ChatGPT to understand what is failing in production. The LLM can filter by project, environment, or specific tags.

"Get the top 10 unresolved issues in the frontend-app project over the last 24 hours. Format them as a table with the issue title, occurrence count, and the number of affected users."

Get Single Issue Details

Tool: get_single_sentry_organization_issue_by_id

Once an LLM identifies a problematic issue, it needs deep context. This tool returns the complete metadata for a specific issue, including the first seen date, last seen date, priority, assignment, and a summary of the latest event culprit.

"Pull the full details for issue ID '123456789'. Tell me if it is a handled or unhandled exception, and identify which release version first introduced this problem."

List Events Bound to an Issue

Tool: sentry_organization_issues_list_events

Issues are groupings; Events are the actual occurrences. This tool allows the AI to fetch the specific crashes associated with an issue. By passing the full=true query parameter, ChatGPT can retrieve the raw stack traces needed for root-cause analysis.

"Fetch the latest 5 events for issue '123456789' with full stack traces included. Analyze the traces and pinpoint the exact file and line number causing the crash."

Query Organization Timeseries Data

Tool: list_all_sentry_organization_events_timeseries

This tool queries Sentry's Discover dataset to return time-series metrics. It is critical for identifying latency spikes, error volume surges, or performance regressions over a specific time window.

"Generate a time-series breakdown of 500-level errors across the backend-api project for the last 6 hours, grouped by the transaction name. Identify which transaction spiked the most."

Trigger Seer Autofix

Tool: create_a_sentry_issue_autofix

Sentry includes an AI engine called Seer that can propose code changes to fix errors. ChatGPT can orchestrate this feature by triggering an Autofix run directly from a chat session, delegating the heavy lifting of repository analysis to Sentry's native tooling.

"Trigger an Autofix run for the TypeError issue '123456789'. Monitor the status and let me know when Seer has generated a proposed pull request."

Bulk Update Issues

Tool: sentry_organization_issues_bulk_update

After triage, ChatGPT needs to take action. This tool allows the agent to mutate issue state - marking bugs as resolved, ignoring them, assigning them to specific developers, or updating their priority levels.

"Mark all issues related to the 'legacy-checkout' tag as ignored until they affect more than 100 users, and assign the critical 'database-lock' issue to the backend engineering team."

Note: This is only a curated selection of tools. For the complete tool inventory, request schemas, and response definitions, view the Sentry integration page.

Workflows in Action

Exposing individual tools is just the foundation. The real power of connecting Sentry to ChatGPT via MCP is the ability to chain these tools into autonomous, persona-specific workflows.

Workflow 1: The Midnight Production Spike (DevOps)

A DevOps engineer is paged at 2 AM for a massive spike in error rates. Instead of manually clicking through Sentry dashboards and cross-referencing logs, they ask ChatGPT for a summary.

"We just got a PagerDuty alert for the payment-gateway project. Find the most frequent issues that started in the last hour, get the stack traces for the worst offender, and summarize the root cause."

Execution Steps:

1. ChatGPT calls list_all_sentry_organization_issues with filters for project=payment-gateway and a time range of the last hour, sorted by event count.
2. ChatGPT identifies the top issue (e.g., a StripeRateLimitError) and extracts the ID.
3. ChatGPT calls sentry_organization_issues_list_events with full=true to retrieve the stack traces for that specific issue.
4. ChatGPT parses the JSON-RPC response, analyzes the stack trace, identifies that a recent deployment lacks retry logic, and generates a plain-English summary for the engineer.

sequenceDiagram
    participant User as DevOps Engineer
    participant ChatGPT as ChatGPT
    participant Truto as Truto MCP
    participant Sentry as Sentry API

    User->>ChatGPT: "Investigate the payment-gateway spike"
    ChatGPT->>Truto: call(list_all_sentry_organization_issues)
    Truto->>Sentry: GET /api/0/projects/org/payment-gateway/issues/
    Sentry-->>Truto: Top Issues Payload
    Truto-->>ChatGPT: Parsed Issues List
    ChatGPT->>Truto: call(sentry_organization_issues_list_events)
    Truto->>Sentry: GET /api/0/issues/{id}/events/?full=true
    Sentry-->>Truto: Stack Trace Payload
    Truto-->>ChatGPT: Parsed Stack Traces
    ChatGPT-->>User: "Root cause identified: Missing Stripe retry logic..."

Workflow 2: Automated Triage and Autofix (Lead Engineer)

A Lead Engineer wants to clear out the morning triage queue quickly by letting AI handle the obvious bugs.

"Find any new issues in the frontend-web project tagged with 'TypeError'. For each one, trigger a Seer Autofix run and assign the issue to the frontend team."

Execution Steps:

1. ChatGPT calls list_all_sentry_organization_issues filtered by project=frontend-web, query=is:unresolved tag:TypeError.
2. For each returned issue ID, ChatGPT calls create_a_sentry_issue_autofix to kick off Sentry's internal code-generation process.
3. ChatGPT calls sentry_organization_issues_bulk_update to update the assignedTo attribute for those issues.
4. ChatGPT responds with a table summarizing the affected issues, their assigned team, and the pending Autofix run IDs.

Security and Access Control

Giving an AI agent access to production telemetry data requires strict governance. Truto's MCP architecture provides robust security controls at the token level, ensuring your agents only access what they absolutely need.

• Method Filtering: Limit the MCP server to specific operation types. By configuring methods: ["read"], you guarantee the agent can query events and metrics but cannot resolve issues, delete projects, or mutate alerting rules.
• Tag Filtering: Restrict tool generation to specific API domains. Passing tags: ["issues", "events"] ensures the LLM never sees tools related to billing, user provisioning, or organization settings.
• Expiration Controls: Assign an expires_at timestamp when generating the MCP server. This allows you to provision temporary, time-boxed access for an external contractor or a short-lived automated script.
• Secondary Authentication: By enabling require_api_token_auth: true, the bare MCP URL is no longer sufficient. The connecting client must also supply a valid Truto API token in the Authorization header, preventing unauthorized access if the URL is accidentally logged or leaked.

A factual note on API limits: Truto does not retry, throttle, or apply backoff on rate limit errors. When the upstream Sentry API returns an HTTP 429 response, Truto passes that error directly back to ChatGPT. Truto normalizes the upstream rate limit information into standardized headers (ratelimit-limit, ratelimit-remaining, ratelimit-reset) per the IETF spec. The caller is entirely responsible for managing retry and backoff logic.

Modernizing Incident Response

Connecting Sentry to ChatGPT via a managed MCP server fundamentally shifts how engineering teams handle incident response. Instead of context-switching between dashboards, writing custom PromQL queries, and manually digging through stack traces, you can interrogate your production environment using natural language.

By leveraging Truto, you eliminate the need to write custom JSON-RPC schemas, manage OAuth token lifecycles, or constantly monitor the Sentry API for deprecations. Truto handles the infrastructure, allowing you to focus on building faster, smarter AI-driven workflows.