history-audit

Installation
SKILL.md

History audit

Answer, from evidence: what are the most common agent failure modes on this machine, how often does each model hit them, and which instruction lines would steer away from them? Corrections live in the user's own messages. The corpus is what the user actually pushed back on, not what the agent self-reported.

Deterministic where possible; agents only where judgment is needed.

Installs
6
Repository
ikkeseb/skills
GitHub Stars
2
First Seen
Aug 18, 2026