AI Blocks › Build and ship software
AI agent monitoring planner
Describe your AI agent. Get a plan to trace its steps, costs and errors without logging private data.
AI agent monitoring planner is a free AI skill for the job "Build and ship software". Give the instructions to an AI assistant that supports the required files and tools, then review its output.
Free · No account · MIT licence · Made by Burhan Sebin · Level: advanced
Made by Burhan Sebin. See the original on GitHub; we copied it as it is, with its licence.
What it does
A plan to see what your AI agent is doing: each step, how long it took, what it cost and where it failed. Nobody's private text gets stored. It starts from the questions you need answered, such as which tool causes timeouts or why cost per task rose. Each user attempt becomes one trace, with a step for every model call, search, tool use, handoff and retry. Steps log fixed fields: version, status, timing, tokens and cost. Content is left out by default, and capture needs approval, masking and short retention. The plan adds a few headline numbers, alerts tied to user impact, and a runbook for investigating problems.
You give it
- A description of the agent's workflow
- Questions you want answered
- Your logging or monitoring tools (optional)
- Privacy and retention rules
- Codebase (optional)
You get
- A trace and event layout
- Metrics with clear definitions
- Dashboard and alert specs
- Privacy and retention settings
- An investigation runbook
Use it when
- Your agent sometimes runs slowly and you cannot tell which step is the cause.
- The bill for your agent rose and you need to know why.
- You need audit evidence of what the agent did without keeping customer text.
- Different tools and frameworks log in different ways and you want one standard.
How to use it
- Add it to your AI once: download it and add the folder to Claude, ChatGPT or Codex as a skill.
- Ask for the job in your own words. Your AI opens the skill when it fits.