otherapi_key

Honeybadger

Honeybadger is an error tracking and application performance monitoring service for developers.

Verdict

Honeybadger tracks errors, uptime checks, and deployments for production apps. In Switchy, @mentioning Honeybadger lets your team report exceptions with full context, log deployments after releases, ping check-ins to confirm cron jobs ran, and upload source maps so minified JavaScript errors become readable. Engineers get the most value during incident response and post-deploy verification. You'll need a Honeybadger API key and project ID; the MCP doesn't read existing errors or query historical data—it only writes new telemetry.

Common use cases

  • Report exceptions during incident triage
  • Log deployments after CI pipeline completes
  • Ping check-ins for scheduled task monitoring
  • Upload source maps post-build for stack traces
  • Send custom telemetry events from chat

Integration

Vendor
Honeybadger
Category
other
Auth
API_KEY
Tools
6
Composio slug
honeybadger

Tools

  • Report Check-In

    Tool to report a check-in (ping) to Honeybadger. Use when monitoring scheduled tasks to signal they're running on time.

  • Report Deployment

    Tool to report a new deployment to Honeybadger. Use after a successful release to notify Honeybadger of the deployment.

  • Report Event

    Tool to send custom events to Honeybadger Insights. Use when recording structured NDJSON telemetry.

  • Report Exception

    Tool to report an exception notice to Honeybadger. Use when sending error details (stack trace, context) for diagnostics.

  • Upload File to S3

    Tool to upload a local file to a managed S3 bucket. Use when preparing files for source-map uploads.

  • Upload Source Map

    Tool to upload JavaScript source maps for error de-minification. Use after deploying assets to Honeybadger to attach source maps.

Setup

Setup guide

  1. 11. In Switchy, open your workspace settings and navigate to the Integrations tab. 2. Click 'Add MCP Integration' and select Honeybadger from the list. 3. Log into your Honeybadger account and go to Settings → API Keys to generate a new personal access token with write permissions. 4. Copy the API key and paste it into the Switchy connection form, then add your Honeybadger project ID (found in your project's settings URL). 5. Click 'Connect' and wait for Switchy to confirm the integration is active. 6. Open any Space, type '@Honeybadger report a test check-in for my-cron-job' and send the message. 7. Check your Honeybadger dashboard under Uptime to verify the check-in appears within 30 seconds.

What teammates see: by default, memories from Honeybadger are scoped to the Space (PROJECT visibility) - you can mark any memory PRIVATE or share it ORG-wide.

Works well with

Top models

Compatibility data appears once enough Spaces have used this MCP together with a given model.

How Switchy teams use it

Not enough Spaces yet to publish anonymised usage stats (we require ≥ 50 Spaces per week).

Starter prompts

Report Deployment

@Honeybadger report a deployment for the api-service repository to production environment, revision abc123def, deployed by the CI system.
Open in a Space →

Log Exception

@Honeybadger report an exception: RuntimeError with message 'Database connection timeout' in the checkout controller, include context user_id=456 and request_id=xyz.
Open in a Space →

Confirm Cron Ran

@Honeybadger report a check-in for the nightly-backup-job so we know it completed successfully this morning.
Open in a Space →

Upload Source Map

@Honeybadger upload the source map from ./dist/app.js.map for minified URL https://cdn.example.com/app.js version 2.1.0.
Open in a Space →

Send Custom Event

@Honeybadger report a custom event with JSON: {"event":"checkout_completed","revenue":49.99,"user_tier":"premium"} to track conversion metrics.
Open in a Space →

Example outputs

Illustrative - representative of the model's voice and quality, not literal recordings.

Prompt

@honeybadger report the exception from our payment gateway timeout — stack trace is in the last error log, user context is session ID abc-789.

Output

Exception reported to Honeybadger. Notice ID: 4f3a2b1c. The error appears as a new occurrence under the "Payment Gateway Timeout" fault group in your project dashboard. Stack trace, session ID, and environment context have been attached. You can view the full diagnostic details at honeybadger.io/projects/12345/faults/67890.

Notes

This example shows error reporting with contextual metadata. Requires a valid API key with write permissions. The MCP does not automatically capture stack traces from your runtime — you must provide the exception details in the prompt or reference a log artifact.

Prompt

@honeybadger report a deployment for version 2.4.1 to production, deployed by CI pipeline run #8821.

Output

Deployment recorded. Honeybadger now associates new errors with release 2.4.1 in the production environment. The deployment appears in your project timeline with metadata: version 2.4.1, environment production, deployed_by CI pipeline run #8821. This helps correlate fault spikes with specific releases.

Notes

Deployment tracking is essential for root-cause analysis — it timestamps when code changes hit production. This example demonstrates the write-action workflow. The MCP does not auto-detect deployments; you must explicitly report them via prompt or CI integration.

Prompt

@honeybadger upload the source map for app.min.js from our latest build, then summarize what source-map coverage we have across all recent deployments.

Output

Source map uploaded successfully for app.min.js (revision abc123). Honeybadger will now de-minify stack traces for errors originating from this file. Based on recent deployment records, you have source-map coverage for 4 of 6 JavaScript bundles deployed in the last 30 days. Missing maps: vendor.min.js (v2.3.0) and analytics.min.js (v1.8.2). Recommend uploading those to improve error readability.

Notes

This synthesis example pairs the upload action with AI reasoning over deployment history. Source-map uploads require the file to exist locally or be accessible via URL. The summarization step is illustrative — actual coverage analysis depends on querying Honeybadger's API for deployment metadata, which may require additional tooling.

Use-case deep-dives

Deploy checklist for 3-person dev team

When Honeybadger MCP replaces your deploy Slack ritual

A 3-person dev team shipping twice a week needs a consistent deploy checklist: tag the release, notify error tracking, upload source maps, confirm the cron still runs. The Honeybadger MCP handles three of those steps in one AI prompt—report the deployment, upload the source map, and ping the check-in—so the on-call engineer doesn't forget the source-map step at 6pm on Friday. This works if your team already uses Honeybadger for errors and uptime; if you're on Sentry or Datadog, this MCP won't help. The trade-off: you're trusting the AI to sequence the S3 upload and source-map attachment correctly, which is fine for small bundles under 10MB but gets flaky with large monorepo builds. If your deploy is three commands and you forget one every third release, connect this MCP and let the AI close the loop.

Customer support triage with error context

Pull stack traces into support threads without tab-switching

A 5-person support team fields bug reports in Intercom, then hunts through Honeybadger to find the matching exception and paste the stack trace back into the ticket. The Honeybadger MCP lets the AI pull error details by exception ID or search recent errors by user email, so the support rep asks 'show me the last error for user@example.com' and gets the stack trace, breadcrumbs, and deploy tag in the chat thread. This saves 90 seconds per ticket when the error exists in Honeybadger; it saves zero seconds if the bug isn't throwing exceptions or if your error tool is something else. The boundary: if your support volume is under 50 tickets a week and most are feature requests, not bugs, the time saved doesn't justify the API-key setup. If you're triaging 20+ exception-driven tickets a day, this MCP turns a four-tab workflow into a one-prompt lookup.

Scheduled task monitoring for ops team

When check-in reporting beats writing a custom ping script

A 2-person ops team runs a dozen nightly ETL jobs and wants to know if any skip a run. Honeybadger's check-in feature expects a ping every N hours; the Honeybadger MCP's report-check-in tool lets the AI send that ping as part of a post-job summary prompt, so the same workflow that logs 'ETL finished, 1.2M rows' also confirms the heartbeat. This works if your jobs already write to a shared log or Slack channel that the AI reads—just add 'and ping Honeybadger check-in XYZ' to the prompt template. The trade-off: if your jobs are silent (no log, no notification), you'll need to wire the AI into the job's exit hook, which is more plumbing than just calling Honeybadger's API directly. If your team already has an AI reading job logs and you want to close the monitoring loop without writing a bash script, this MCP is the two-line addition you need.

Frequently asked

What does the Honeybadger MCP do in Switchy?

It lets AI agents report errors, deployments, and check-ins directly to your Honeybadger project without writing integration code. Agents can send exception traces with context, notify Honeybadger when a release ships, ping scheduled task monitors, and upload source maps for JavaScript error de-minification. Useful when you want AI to handle incident reporting or deployment notifications as part of a workflow.

Do I need admin access to connect Honeybadger?

You need a Honeybadger API key with write permissions for the project you're connecting. Standard project members can generate this from their Honeybadger settings. The key grants access to report errors, deployments, and events — it doesn't expose sensitive project data or billing controls. One API key works for the entire Switchy workspace; you don't need separate keys per user.

Can the MCP read existing errors or search past deployments?

No. The six tools are write-only: they send data to Honeybadger but can't query it. If you need an agent to check error trends or retrieve deployment history, you'll have to use Honeybadger's REST API directly or export data to another tool. This MCP is for pushing telemetry into Honeybadger, not pulling reports out.

Why use this instead of calling Honeybadger's API from code?

You skip the boilerplate. Instead of writing HTTP clients and handling auth in every script, you describe what happened in natural language and the agent picks the right tool. The source map upload flow is especially tedious to script manually — this MCP handles S3 pre-signing and the two-step upload for you. Trade-off: you lose fine-grained control over request headers and retry logic.

Who on the team should connect this MCP?

Whoever owns deployment pipelines or on-call runbooks. If your team uses AI agents to automate releases or respond to incidents, that person should add the Honeybadger connection. The API key they provide will be shared across all workspace members, so coordinate with your Honeybadger admin to ensure the key has appropriate project scope and doesn't expire mid-quarter.

Data last verified 607 hours ago.Sources aggregated hourly to weekly. See docs/architecture/model-directory.md.