An open-source, coding agent harness designed specifically for local LLMs. Designed as everything a user needs to jumpstart their own harness (detailed event steams, logs, and performance data) or quickly replace closed source one (featuring out of the box plugins for web search, image gen, memory, auto research and more in dev).
Skill Claude CodeCodex
Creates or improves a LocalHarness skill, a reusable procedure in a SKILL.md. Use when the person says make a skill, turn this into a skill, new skill, save this as a procedure or improve this skill.
Skill Claude CodeCodex
Writes a failing test that reproduces a reported bug before any fix is made. Use when asked to reproduce a bug, add a regression test or prove a bug with a test.
Skill Claude CodeCodex
Writes a commit message for the staged changes: a component prefix, a colon and one lowercase sentence saying what changed. Use when asked to write, draft, reword or fix a commit message.
Skill Claude CodeCodex
Adds or tightens one-line docstrings across a module without changing its code. Use when asked to document functions, add docstrings or clean up the docstrings in a file.
Skill Claude CodeCodex
Writes a short incident summary with the timeline, the impact, the cause and the follow-ups. Use when asked for a postmortem, an incident write-up or a summary of an outage.
Skill Claude CodeCodex
Drafts user-facing release notes from the changes since the last tag, grouped into features, fixes and breaking changes. Use when asked for release notes, a release announcement or what is new in a version.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: