Back to directory

oh-my-agent

oh-my-agent is a skill pack and harness for AI coding agents that verifies their work with stop-hook test gates, artifact checks, and an independent judge, instead of trusting the agent's own summary.

oh-my-agent

oh-my-agent makes an AI coding agent's claim of finished work falsifiable. A Stop hook blocks the end of a session until the project's own typecheck, test, and lint commands exit cleanly, and a gate command checks for the artifacts a workflow must have produced rather than accepting the agent's narrated summary.

An independent judge with a fresh context re-verifies every completion criterion each round, including ones that already passed, and every gate decision is written to an append-only event log that can be reviewed later. The project can be installed as a lighter skill pack for Claude Code, Cursor, Codex and other agent runtimes, or as a full harness with workflows, rules, keyword-detection hooks, and an agent-spawning command, applying the same checks across about a dozen agent runtimes from one .agents/ directory.

It is written in TypeScript and installs with npx or a shell/PowerShell install script; the full harness also installs bun, uv, and serena.

More in Agent skills

See all 581
  • Provenance gate

    github.com

    A deterministic authorization gateway that sits between an AI agent and its tools, tracing where each parameter of a side-effect call came from, and blocking the call or asking a human when the source does not check out.

  • A markdown framework of system rules that restates Isaac Asimov's Four Laws of Robotics as a set of prioritized loss calculations for large-language-model agents to follow, published as a discovery file for agents and crawlers to read.

  • Vitruvius

    github.com

    An engineering research agent that runs a discover, read, synthesize, verify and review loop over technical questions, reading sources directly rather than inferring from titles, and publishing its own scored verifier benchmark, weaknesses included.