uinaf/autoreview
Run structured Codex/Claude autoreview closeout: choose the target, collect schema-validated findings, and rerun tests plus review until clean.
72
90%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Passed
No known issues
upstream.mdreferences/
Upstream
This skill tracks OpenClaw's shared agent-skills repo as its canonical upstream.
- Source: https://github.com/openclaw/agent-skills/tree/main/skills/autoreview
- Applied upstream commit:
3446a70
This repo's version keeps uinaf/agents packaging files such as tile.json,
agents/openai.yaml, and this provenance note around the upstream skill core.
It also prunes engine/platform branches Altay does not use: the local helper is
Codex + Claude Code only, with no Droid, Copilot, or PowerShell harness wrapper.
When borrowing future updates, prefer copying OpenClaw's SKILL.md and bundled
helper scripts first, then reapply only this repo's packaging and pruning.