Improve agent harness configuration based on recent friction points. Use when invoked as `$retro` or when the user asks for autonomous cleanup of Codex, agent harness, or migrated Claude configuration.
87%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Passed
No findings from the security scan
Apply targeted agent-configuration improvements based on recent friction points.
Use the user's supplied friction points. If none are provided, ask for 2-5 concrete issues from recent sessions before editing.
Prefer active Codex and shared harness paths:
.codex/.agents/AGENTS.md or provided global/project instructionsRead or edit .claude/ only when the friction point is explicitly about Claude parity or a migration source.
Load relevant baseline instructions and config.
For each friction point, identify root cause:
Create a concrete fix plan listing file, change, and prevented failure.
Apply only targeted edits. Preserve unrelated config.
Validate modified files:
Report:
## Retrospective Report
### Friction Points
- <issue> → <root cause> → <fix>
### Changes Made
| File | Change | Prevents |
|---|---|---|
### Validation
- ...phatblat/dotfiles
4d65cab
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.