Laravel Telescope MCP exposes Telescope telemetry data via the Model Context Protocol (MCP), enabling AI assistants to access application metrics such as logs, queries, requests, and exceptions. This helps you get quick, precise insights from tools like Cursor, Claude, or Copilot Chat without manual data digging.

How to use

To use this MCP server, install it in your Laravel application, start the MCP endpoint, and connect your AI assistant client to the provided URL. Once connected, you can query a wide range of Telescope telemetry data through natural language or direct MCP tool commands. Your AI assistant will fetch data like recent HTTP requests, slow queries, exceptions, logs, and more, and present results in readable summaries alongside structured data.

How to install

composer require lucianotonet/laravel-telescope-mcp

php artisan vendor:publish --provider="LucianoTonet\\TelescopeMcp\\TelescopeMcpServiceProvider"

# Optional: enable MCP in your environment
#TELESCOPE_MCP_ENABLED=true
#TELESCOPE_MCP_PATH=telescope-mcp

Then start your Laravel server as you normally would (e.g., via php artisan serve or your preferred web server). The MCP endpoint will be available at the path you configured, typically something like http://your-app.test/telescope-mcp/manifest.json. When you access the manifest URL in a browser, you should see MCP metadata describing the available tools and data sources.

## Configuration and usage notes

Environment and access: Use the optional environment variables to enable and locate the MCP endpoint. You can customize the endpoint path with TELESCOPE\_MCP\_PATH. The endpoint URL will combine your APP\_URL with this path.

Security: Protect the MCP endpoint with appropriate middleware (for example, auth:sanctum or auth.basic) to ensure only authorized clients can fetch data.

## Troubleshooting

Connection problems often come from network or URL mismatches. Ensure your MCP URL is reachable from the client and use the exact path shown by the server configuration.

If you see empty results for a tool, verify that Telescope is recording data for that data type and that you are querying with valid parameters. Some tools may require specific flags or filters to return data.

## Quick Start

composer require lucianotonet/laravel-telescope-mcp

php artisan vendor:publish --provider="LucianoTonet\TelescopeMcp\TelescopeMcpServiceProvider"

Connect with your AI assistant using the following MCP URL

For example, set the base URL to your app and the path to telescope-mcp

http://127.0.0.1:8000/telescope-mcp

Examples of direct MCP tool usage from your AI assistant

@laravel-telescope-mcp requests --limit 5 @laravel-telescope-mcp exceptions --limit 3 @laravel-telescope-mcp queries --slow true --limit 10

You can also query by natural language, and your AI assistant will translate that into the appropriate MCP tool calls.

Usage Examples

Direct MCP tool usage from your AI assistant after connecting to the MCP endpoint: you can list recent entries, drill into specific items by ID, and apply filters such as time ranges or severity levels.

Current status and notes

This MCP server provides integrated access to Telescope telemetry through MCP, enabling AI assistants to fetch and summarize data with structured responses and human-readable outputs.

Security and operation details

Authentication and access control are recommended for the MCP endpoint. You can enable middleware like auth:sanctum or basic authentication to restrict usage. Adjust timeouts and payload limits in your configuration to fit your hosting environment.

Available tools

requests

Records incoming HTTP requests with details such as id, method, status, and path

exceptions

Tracks application errors with stack traces and context

queries

Monitors database queries with performance metrics, including slow queries

logs

Records application logs with level and message details

http_client

Monitors outgoing HTTP requests and responses

mail

Monitors email sending activity with recipient and subject details

notifications

Records notification dispatches including channel and status

jobs

Tracks queued job executions with status and queue information

events

Monitors event dispatches within the application

models

Tracks Eloquent model operations such as create, update, delete, and retrieve

cache

Monitors cache operations with keys and results

redis

Tracks Redis commands and their effects

schedule

Monitors scheduled task executions and outcomes

views

Records view renders for troubleshooting UI performance

dumps

Records var_dump and dd() usage with location data

commands

Tracks Artisan command executions and outcomes

gates

Records authorization checks and results

batches

Lists and analyzes batch operations, showing status and names

prune

Removes old Telescope entries (potentially needs attention)