Skip to content
Get packs

Convex app log and health queries

Ask an agent plain-language questions about a running Convex app's errors, slow functions and deploy impact.

Source Download skill
Repository
openclaw/clawhub
Stars
9,546
Licence
MIT
Files
1 file

What it does

Teaches an AI agent how to read the logs and health data of a live Convex app (a hosted backend platform) using the official Convex MCP tools. When you ask something like 'what failed after the last deploy?', the agent fetches a recent window of logs, filters and counts the failures locally, and answers with a one-line finding, supporting evidence and a link to the Convex dashboard so you can verify it yourself. Performance and cost causes are handed off to a separate advisor skill rather than fixed here.

Great for

  • Developers running a Convex backend who want quick answers about recent errors without digging through the dashboard
  • Teams checking whether a recent deploy caused new failures
  • People investigating why one specific request or function call failed

Skip it if

  • Apps not built on Convex, since the skill only works with the Convex MCP tools (not useful outside that platform)
  • Fixing slow or expensive functions directly, because the skill only reports those and defers the fix to convex-advisor
  • Production debugging where you want the agent to change data or settings, since the skill is designed to be read-only but still depends on the correct deployment being chosen

Try asking

  • “What functions are throwing errors in my dev deployment in the last few hundred log lines?”
  • “Did my latest deploy break anything? Compare failure timing with the deployment version.”
  • “Why did the call to sendMessage fail for requestId abc123?”
  • “Show me any read-limit or write-conflict events from the last 72 hours on prod.”
Step-by-step guide: how to use Convex app log and health queriesInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the MIT licence and an attribution note.

    convex-insights.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip convex-insights.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip convex-insights.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Name the deployment and function you care about up front so the agent scopes its query instead of pulling broad logs.
  • Ask one question at a time; the skill picks a single view (failures, health or trace) per question.
  • Use the dashboard link in the answer to confirm findings before acting on them.

What's inside

  • SKILL.md

Redistributed unmodified from openclaw/clawhub under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.