Back to directory

Claude Code Harness

A Claude Code and Codex plugin that delegates planning, implementation and review to the agent through a disciplined plan, work, review, sync and release loop, with the developer approving each stage.

Claude Code Harness

Claude Code Harness is a development plugin that delegates planning, implementation, validation and review to Claude Code or Codex. A developer states the intended outcome and completion criteria, and the assigned agents inspect the existing code, organize the work, and return changes, verification results and unknowns for the developer to judge against those criteria; the agent chooses its method, but the developer decides what to build, what counts as complete, and which actions are authorized.

Five verb skills keep the process bounded: plan turns intent into a scope and acceptance-criteria document that the developer approves or corrects before anything runs; work executes the approved scope, choosing solo or team execution by task count and applying test-driven development where the task requires it; review checks the result separately from implementation, with major findings blocking completion; sync compares the plan against what was actually built and reports drift; and release packages only verified evidence into a changelog, tag and release.

It installs through Claude Code's plugin marketplace command and is written in Shell under the MIT license.

More in Coding agents

See all 793
  • The City

    github.com

    A native macOS app that shows Claude Code runs as a city, with each project as a building, each team as a floor of robots, and every animation driven by a real event from the claude CLI's stream.

  • rig

    tryrig.ai

    A small, single-binary agent runtime that points at any OpenAI-compatible model, local or hosted, reads and edits a repository's code, runs its tests and remembers what it learned, with a loop implemented in a few hundred lines.

  • Jauvex

    jauvex.reindent.com

    An Electron desktop app that puts Claude Code, Codex and Grok coding-agent sessions side by side and lets you steer them by voice, with a small model handling fast routing decisions.