mcpField
C
convex-sentinel

convex-sentinel

github.com/get-convexCloud & DevOps6931.0k installsUpdated 1d ago
Repository

About convex-sentinel

convex-sentinel is a skill for recording Convex production errors in the user's own deployment, including server function failures, client crashes, and concurrency or scale signals. It redacts data by default at write time and can react to new errors or hand them to an optional workflow.

Key features

  1. 01Captures server and client errors
  2. 02Records concurrency and scale signals
  3. 03Redacts data by default at write time
  4. 04Views errors through the Convex CLI
  5. 05Reacts to new errors via an event
  6. 06Optionally enables a self-healing cron

Use cases

  1. 01Review production function failures in one place
  2. 02Capture React crashes and unhandled browser errors
  3. 03Start a fix-PR flow for recurring errors

MCP support

New errors can trigger the monitor's prod_error event. The README also describes querying error records with the Convex CLI, but does not specify an MCP interface.

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/get-convex/agent-skills --skill convex-sentinel

After installing, type/convex-sentinelin Claude Code to confirm it loaded.

Details

From the project README

Capture production errors in your own deployment

Install @convex-dev/sentinel to capture production errors (server function failures, client JS/React crashes, OCC and scale signals) into a table in the user's OWN deployment, redacted at write time, then react to new ones. Data never leaves the user's deployment.

Workflow

  1. Install the component: app.use(sentinel) in convex/convex.config.ts.
  2. Wire the client SDK: a React error boundary plus window.onerror/unhandledrejection and breadcrumbs.
  3. Redaction runs at write time and is on by default (default-deny on secret key names and value patterns).
  4. Read recent errors with the Convex CLI (convex data, run-once-query); react to new ones via the monitor's prod_error event.
  5. Optionally enable the self-healing cron: triage classifies each error and, for recurring non-transient ones, hands it to ai-runner to open a fix PR.

Rules

  • Redaction is mandatory and on by default — never store raw secrets; the agent's reads reach the model provider.
  • Data stays in the user's deployment; never send it to a third party.
  • Sample and cap to control volume and cost.
  • Capturing PROD errors needs a deployed cloud app; install works anonymously.

Comments

Markdown supported