Create, convert, audit, and package executable AI agent workflow patterns as Codex/Claude-style skills. Use when the user wants an agent workflow library, reusable workflow schema, pattern-to-skill conversion, skill pack design, exporter/validator behavior,…
Mine local Codex and Claude chat/session records into reusable agent workflow candidates. Use when the user asks to inspect agent chat history, find repeatable workflows, decide what should become a skill, compare new-skill candidates against existing skills,…
Create compact, evidence-backed handoff packs for long-running agent tasks. Use before context compaction, Codex/Claude session handoff, task resumption, or when the user asks for exact current status across files, constraints, tests, decisions, blockers,…
Force an explicit review step before any agent-generated diff is applied, committed, or landed. Collect diff, run risk analysis, produce a review pack, and require human approval of the plan and risks. Use before git commit, push, PR creation, or direct…
Create, audit, and harden skills so they become self-maintaining instead of cat-and-mouse. Use when designing new skills, reviewing existing ones for drift, or turning brittle workflows into reliable ones. Forces Negative Examples, Verification Checkpoints,…
Diagnose failing CI by collecting fresh check evidence, reproducing the smallest failing command locally when possible, identifying one defensible root cause, and proposing a minimal fix plan. Use when CI is red, tests differ between local and remote, or the…