gfg-sentry-mcp-server 0.1.0

MCP server for querying Sentry alerts and issues.

Sentry MCP Server (Python)

This project provides a Python MCP server with tools for:

• list alerts
• retrieve alert detail (including rule payload)
• search issues
• retrieve an issue
• list issue events
• retrieve issue event

Tools

1. sentry_list_alert_rules
   • Lists issue alert rules and/or metric alert rules.
2. sentry_retrieve_alert_rule_detail
   • Retrieves a single alert rule detail (metric or issue).
3. sentry_search_issues
   • Searches issues using Sentry query syntax.
   • Parameters:
     • query (required): Sentry issue search query.
     • org_slug (optional): overrides SENTRY_ORG_SLUG.
     • limit (optional): max results, clamped to 1..100 (default 25).
   • Returns:
     • organization: organization slug used.
     • query: original query string.
     • count: number of issues in this response.
     • issues: raw issue objects from Sentry.
   • Notes:
     • This tool fetches one page of results only.
     • For large result sets, use narrower filters and run multiple queries.
4. sentry_retrieve_issue
   • Retrieves one issue by issue ID.
5. sentry_list_issue_events
   • Lists error events for a specific issue.
   • Parameters:
     • issue_id (required): global Sentry issue ID.
     • org_slug (optional): overrides SENTRY_ORG_SLUG.
     • start / end (optional): ISO-8601 time range.
     • stats_period (optional): relative range like 24h, 7d (overrides start/end).
     • environment (optional): one or more environment filters.
     • full (optional): include full event body payload (stacktrace/context).
     • sample (optional): deterministic pseudo-random event order.
     • query (optional): Sentry event search query.
     • cursor (optional): pagination cursor from Sentry response headers.
   • Returns:
     • organization: organization slug used.
     • issue_id: issue ID used.
     • count: number of events in this response.
     • events: raw event objects from Sentry.
6. sentry_retrieve_issue_event
   • Retrieves one specific issue event.
   • Parameters:
     • issue_id (required): global Sentry issue ID.
     • event_id (required): event ID, or latest, oldest, recommended.
     • org_slug (optional): overrides SENTRY_ORG_SLUG.
     • environment (optional): one or more environment filters.
   • Returns:
     • organization: organization slug used.
     • issue_id: issue ID used.
     • event_id: event identifier used.
     • event: raw event object from Sentry.

search_issues query examples

Use Sentry issue search filters in the query field:

• is:unresolved level:error
• environment:production release:1.2.3
• user.email:[e-mail]
• transaction:/api/orders has:stacktrace
• project:backend assigned:me

Required Environment Variables

• SENTRY_AUTH_TOKEN: Sentry token with sufficient read permissions.
• SENTRY_ORG_SLUG: Default organization slug (optional if always passed to tools).
• SENTRY_BASE_URL: Optional, defaults to https://sentry.io/api/0.

Install

python -m venv .venv
source .venv/bin/activate
pip install -e .

Run

export SENTRY_AUTH_TOKEN="YOUR_TOKEN"
export SENTRY_ORG_SLUG="your-org"
sentry-mcp-server

Example MCP config snippet

Use this in your MCP client configuration:

{
  "mcpServers": {
    "sentry": {
      "command": "sentry-mcp-server",
      "env": {
        "SENTRY_AUTH_TOKEN": "YOUR_TOKEN",
        "SENTRY_ORG_SLUG": "your-org"
      }
    }
  }
}

Publish to PyPI

This repository includes .github/workflows/publish-pypi.yml for trusted publishing.

1. Create a release tag like v0.1.1.
2. Push the tag to GitHub.
3. GitHub Actions builds the package and publishes it to PyPI.