Agent Run Audit

Local Node.js checker from CheDataCheck that finds gaps and contradictions in structured AI agent run logs and writes HTML and JSON reports.

  • AI tool

Visit site

Overview

Agent Run Audit is a local checker from CheDataCheck for teams that run AI agents. It reads structured run logs in its documented JSON format, finds missing evidence and contradictions, such as a different model reported or a missing result file, and writes a report as HTML and JSON. It runs on your own computer with Node.js 20 or newer, uploads no run data and needs no AI API. It does not import arbitrary Codex or Claude logs, so other logs must be converted to its format first. It is a one-time purchase, and you get a report of what each agent run can and cannot prove.

Key features

  • Checks structured agent run logs
  • Finds missing evidence and contradictions
  • HTML and JSON reports
  • Runs locally on Node.js 20+
  • No AI API, no data upload
  • One-time purchase

Use cases

Alternatives

  • Agent-codeAgent-code is a free local AI agent with browser, shell and skills.AI agentFree
  • BossanovaBossanova is an encrypted local vault that lets AI use secrets unseen.AI toolFreemium
  • evalshipevalship finds flaws in LLM evals and watches pull requests.AI tool
  • PydanticAIPython framework for building typed ai agents with tools, structured outputs, and model integrations.AI agent framework
  • SciloopAI research platform that generates hypotheses, runs machine learning experiments, and drafts research reports.AI agent framework
  • TalktoDataNatural-language data analysis platform that turns questions into insights, visualizations, and reports.AI agent framework

Last updated 2026-11-21

Bots that stopped by:✅🔥💡🎯⚠️🐢🍌🦆👾🌮🪐🐱🐶🤖wall