Skip to content

Connect ExecVision to Claude: Manage Coaching & Sales Performance

Learn how to build and configure a managed MCP server to connect Claude to ExecVision, enabling AI agents to analyze call transcripts and coaching plans.

Nidhi KN Nidhi KN · · 8 min read
Connect ExecVision to Claude: Manage Coaching & Sales Performance

If you need to connect ExecVision to Claude to automate sales coaching audits, analyze call transcripts, or track team KPIs, you need a Model Context Protocol (MCP) server. This server acts as the translation layer between Claude's tool calls and ExecVision's REST APIs. You can either build and maintain this infrastructure yourself, or use a managed integration platform like Truto to dynamically generate a secure, authenticated MCP server URL.

If your team uses ChatGPT, check out our guide on /connect-execvision-to-chatgpt-analyze-call-data-transcripts/ or explore our broader architectural overview on /connect-execvision-to-ai-agents-audit-call-reviews-team-activity/.

Giving a Large Language Model (LLM) read and write access to a sprawling conversation intelligence platform like ExecVision is an engineering challenge. You have to handle authentication token lifecycles, map massive JSON schemas to MCP tool definitions, and deal with ExecVision's strict API patterns. Every time ExecVision updates an endpoint or deprecates a data model, you have to update your server code, redeploy, and test the integration.

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

The Engineering Reality of the ExecVision API

A custom MCP server is a self-hosted integration layer. While the open MCP standard provides a predictable way for models to discover tools, the reality of implementing it against ExecVision's APIs is painful. You are not just integrating a standard REST API - you are navigating specific vendor constraints designed for enterprise compliance.

If you decide to build a custom MCP server for ExecVision, you own the entire API lifecycle. Here are the specific challenges you will face:

Complex Pagination Mechanics ExecVision does not use standard offset or cursor-based pagination. Endpoints like calls.list require a combination of since (the timestamp of the last change) and last_id (the UUID of the last call processed). If you just pass a page number, the request fails. An LLM has no context on how to construct this composite pagination key. Your MCP server must explicitly guide the model to extract and return these exact values in subsequent calls.

Strict Search Syntax The exec_vision_calls_search endpoint is notoriously rigid. You can only search by exactly one metadata field at a time (e.g., metadata.company). Furthermore, the value must be perfectly lowercased, and spaces must be replaced by the + character. If the LLM sends "Hewlett Packard" instead of "hewlett+packard", the search will silently return empty or throw a validation error.

Vendor-Gated Write Operations Operations like deleting a call via the API (delete_a_exec_vision_call_by_id) are not enabled by default, regardless of your OAuth scopes or API key privileges. Call deletion must be explicitly enabled for your API user by Mediafly (ExecVision's parent company) on the backend. If an LLM attempts to delete a call without this flag, it receives a 403 Forbidden error, which can cause agent loops if the model repeatedly tries to "fix" its payload.

Generating a Managed ExecVision MCP Server

Instead of building a Node.js or Python server from scratch, you can use Truto to generate a hosted MCP server. Truto dynamically derives the tool schemas from the ExecVision integration definitions, mapping query and body parameters into a flat namespace that Claude understands.

You can create the server in two ways: via the Truto UI or programmatically via the API.

Method 1: Via the Truto UI

For ad-hoc agent testing or internal administrative workflows, the UI is the fastest path.

  1. Log into your Truto dashboard and navigate to the integrated account page for your ExecVision connection.
  2. Click the MCP Servers tab.
  3. Click Create MCP Server.
  4. Select the desired configuration (name, allowed methods, tags, and an optional expiry time).
  5. Copy the generated MCP server URL (e.g., https://api.truto.one/mcp/a1b2c3d4e5f6...). This URL contains a cryptographically hashed token that authenticates requests to this specific tenant.

Method 2: Via the Truto API

For production applications, you will want to generate MCP servers programmatically on behalf of your users.

Make a POST request to /integrated-account/:id/mcp:

curl -X POST https://api.truto.one/integrated-account/<execvision_account_id>/mcp \
  -H "Authorization: Bearer <YOUR_TRUTO_API_TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Claude ExecVision Server",
    "config": {
      "methods": ["read"],
      "require_api_token_auth": false
    }
  }'

The response will contain the secure URL you need to provide to Claude:

{
  "id": "mcp_8a9b0c1d2e",
  "name": "Claude ExecVision Server",
  "config": { "methods": ["read"] },
  "expires_at": null,
  "url": "https://api.truto.one/mcp/a1b2c3d4e5f6..."
}

Connecting the MCP Server to Claude

Once you have the URL, you need to configure Claude to consume it. Truto's MCP servers communicate via JSON-RPC over Server-Sent Events (SSE).

Via the Claude UI (Claude for Enterprise/Teams)

If your organization supports custom connectors via the UI:

  1. In Claude, navigate to Settings -> Integrations -> Add MCP Server (or Custom Connectors).
  2. Name the integration (e.g., "ExecVision by Truto").
  3. Paste the Truto MCP URL.
  4. Click Add. Claude will instantly handshake with the server and list the available ExecVision tools.

Via Manual Configuration File (Claude Desktop)

For local development or standard Claude Desktop installations, update your claude_desktop_config.json file. Since Truto uses SSE, you will use the official @modelcontextprotocol/server-sse package to wrap the HTTP endpoint.

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

Restart Claude Desktop. The "plug" icon will appear, indicating that the ExecVision tools are now loaded into the model's context window.

ExecVision Hero Tools

Truto exposes dozens of ExecVision resources as discrete, highly documented MCP tools. Here are the highest-leverage tools for coaching and sales workflows.

This tool allows the agent to find specific call IDs based on CRM metadata (like company name or deal ID). Because of ExecVision's strict search syntax, the Truto schema explicitly instructs the LLM to provide exactly one field and format the value correctly.

"Find all calls associated with the company 'Acme Corp'. Search metadata.company for acme+corp."

get_single_exec_vision_call_by_id

Retrieves the complete dataset for a specific call, including CRM matches, call stats, metadata, and the transcript. This is the primary tool for deep-dive call analysis.

"Fetch the full details and transcript for call ID 123e4567-e89b-12d3-a456-426614174000 and summarize the key objections raised by the prospect."

list_all_exec_vision_call_topics

Lists topic (tracker) matches across completed calls. This is incredibly useful for tracking macro trends, like how often competitors or specific product features are mentioned.

"List all call topics discussed since last Monday where the phrase 'pricing' was mentioned at least 3 times."

list_all_exec_vision_coaching_plans

Retrieves active coaching plans, including objectives, coach and recipient IDs, planned meetings, and target skills. This tool enables agents to audit management activity.

"Get the coaching plans modified this month and tell me which reps have missed their planned meetings with their managers."

list_all_exec_vision_scores

Fetches scorecard answers from call reviews. This includes raw answers, normalized percentages, and the specific skills being graded, allowing the agent to identify coaching gaps.

"Pull the scorecard answers for the last 30 days and calculate the average score for the 'Discovery Questions' skill."

delete_a_exec_vision_call_by_id

Deletes a specific call record. Note that this tool will return a 403 error unless Mediafly has explicitly enabled call deletion for your API user.

"Delete the call with ID 987fcdeb-51a2-43d7-9012-3456789abcde. If you receive a permission error, let me know that I need to contact Mediafly support."

For the complete inventory of available endpoints, schemas, and required parameters, refer to the ExecVision integration page.

Workflows in Action

By chaining these tools together, Claude can execute multi-step analytical and administrative tasks that normally require hours of manual work inside the ExecVision dashboard.

Workflow 1: Sales Coaching Audit

A VP of Sales wants to understand why a specific rep is struggling with a new product pitch.

"Find the recent calls for John Doe where the topic 'New Product Alpha' was discussed. Pull the transcripts for those calls, compare them against John's recent scorecard answers for 'Product Knowledge', and identify where he is missing the key value propositions."

  1. Claude calls list_all_exec_vision_call_topics to find calls matching the user ID and topic.
  2. Claude loops through the returned IDs, calling get_single_exec_vision_call_by_id to retrieve the transcripts.
  3. Claude calls list_all_exec_vision_scores filtered by the rep's ID to retrieve recent evaluations.
  4. Claude analyzes the transcripts against the scorecards, providing a detailed summary of the rep's missed talking points and correlating them with poor review scores.
sequenceDiagram
    participant C as Claude
    participant MCP as Truto MCP Server
    participant EV as "Upstream API (ExecVision)"
    
    C->>MCP: call list_all_exec_vision_call_topics
    MCP->>EV: GET /api/v1/topics
    EV-->>MCP: Topics & Call IDs
    MCP-->>C: JSON list
    
    C->>MCP: call get_single_exec_vision_call_by_id<br>(for each ID)
    MCP->>EV: GET /api/v1/calls/{id}
    EV-->>MCP: Transcript Payload
    MCP-->>C: Transcript Data
    
    C->>MCP: call list_all_exec_vision_scores
    MCP->>EV: GET /api/v1/scores
    EV-->>MCP: Scorecard Data
    MCP-->>C: Scores JSON

Workflow 2: Competitor Trend Analysis

A Product Marketing Manager needs to know how often a specific competitor is disrupting deals.

"Search for calls over the last 14 days where the company 'Competitor X' was mentioned. Pull those transcripts and summarize the specific reasons prospects gave for considering them over us."

  1. Claude calls exec_vision_calls_search formatted perfectly for the API (e.g., competitor+x).
  2. Claude extracts the call IDs and invokes get_single_exec_vision_call_by_id to get the transcripts.
  3. Claude reads the transcripts, identifies the context around the competitor mentions, and synthesizes a report on the competitive landscape.

Security and Access Control

Exposing conversational data to an LLM requires strict governance. Truto MCP servers provide multiple layers of access control configured at the server generation stage:

  • Method Filtering: Use methods: ["read"] to ensure the MCP server can only execute get and list operations. This strictly prevents the LLM from accidentally deleting calls or altering coaching plans.
  • Tag Filtering: Restrict tools to specific functional areas using tags. You can scope a server so it only sees tools related to scorecards and coaching, completely hiding transcripts if privacy requires it.
  • Expiration (expires_at): Create short-lived MCP servers for temporary investigations. Truto automatically cleans up the server and invalidates the token when the timestamp is reached.
  • API Token Enforcement: Enable require_api_token_auth to force the MCP client to pass a valid Truto API token in the headers, adding an extra authentication layer beyond the secure URL.

Understanding Rate Limits and Error Handling

When connecting AI agents to third-party APIs, rate limiting is a critical architectural consideration. LLMs are fast and can easily trigger 429 Too Many Requests errors if they try to pull transcripts in a tight loop.

Factual note on rate limits: Truto does not retry, throttle, or apply backoff on rate limit errors. When the upstream ExecVision API returns an HTTP 429, Truto passes that error directly back to the caller.

Truto normalizes the upstream rate limit information into standardized IETF headers (ratelimit-limit, ratelimit-remaining, ratelimit-reset). The caller (or the agent framework) is completely responsible for reading these headers, pausing execution, and implementing retry and backoff logic. Do not expect the MCP server to magically absorb rate limit errors.

Moving Forward

Connecting ExecVision to Claude via MCP transforms conversational intelligence from a static dashboard into a dynamic, queryable dataset. By utilizing managed MCP servers, you bypass the friction of OAuth maintenance, custom pagination logic, and strict search syntax formatting.

Your engineers do not have to write custom integration code, and your sales managers gain a conversational interface for auditing coaching plans, analyzing lost deals, and identifying systemic team weaknesses.

FAQ

Does Truto handle ExecVision API rate limits automatically?
No. Truto passes HTTP 429 rate limit errors directly to the caller and standardizes the rate limit headers (ratelimit-limit, ratelimit-remaining, ratelimit-reset). The caller is responsible for implementing retry and backoff logic.
Can I prevent Claude from deleting ExecVision calls?
Yes. When generating the MCP server in Truto, you can set the config to only allow 'read' methods, which prevents write operations like delete. Furthermore, call deletion requires explicit permission from Mediafly on the ExecVision backend.
How does the MCP server handle ExecVision's complex search endpoint?
Truto dynamically generates schemas that explicitly instruct the LLM on how to format requests, such as lowercasing strings and replacing spaces with '+' for the exec_vision_calls_search endpoint.

More from our Blog