audit
01Command Claude Code
Self-critique memory files — flag drift, propose edits (no auto-execute).
Drop-in thinking-partner personality layer for Claude Code, Cursor, ChatGPT, and 20+ AI agent runtimes. MIT. Plaintext. Persistent memory. Framework-grounded. Calibration-disciplined.
Command Claude Code
Self-critique memory files — flag drift, propose edits (no auto-execute).
Command Claude Code
Log a prediction with probability and resolve-by date.
Command Claude Code
Memory-wiring diagnostic — detect a host auto-memory shadow layer, fact-class overlap, and absolute paths.
Command Claude Code
Structured decision analysis through the gate chain.
Command Claude Code
Gate 2 (Assent) on demand — interrogate a claim.
Command Claude Code
Cluster instincts and propose graduations to skills, knowledge, or rules.
Command Claude Code
Graceful memory deletion with audit trail.
Command Claude Code
Explicitly invoke a named framework from KNOWLEDGE.md on the current problem.
Command Claude Code
Domain learning mode — the user teaches you something.
Command Claude Code
Imagine the project failed in 6 months — what broke, what we missed.
Command Claude Code
Promote a project-scope instinct to global scope.
Command Claude Code
End-of-session reflection — summarize, update memory, note open questions.
Command Claude Code
Review pending predictions and resolve any past their date.
Command Claude Code
Quick state check — memory files, predictions, Brier score, last session.
Command Claude Code
Construct the strongest version of a position before disagreeing.
Command Claude Code
Drain a host/buffer memory layer into Wick's canonical memory/ folder, with consent.