What problem does it solve? Codebases contain hidden fragilities—implicit assumptions, shared mutable state, invisible invariants—that only surface when a future developer makes a reasonable edit. This Skill identifies those fragile patterns before they break by writing realistic incident reports for bugs that haven't happened yet. ## Core Features & Use Cases - Fragility Detection: Reads code deeply and matches it against an 11-pattern catalogue covering ordering dependencies, stringly-typed contracts, coincidental correctness, security fragility, and more. - Fictional Post-Mortems: Produces concrete, past-tense incident reports in ANTE-MORTEM.md naming actual functions, variables, and plausible future edits. - Hardening TODOs: Extracts actionable suggestions into TODOS.md tagged as [CHORE], [INVESTIGATE], [SECURITY], or [BUG] for real defects found during analysis. - Use Case: Before a major refactor, run the Skill on a core module to surface architectural assumptions that could silently break, then hand the generated TODOs to the team as a hardening backlog. ## Quick Start Ask the AI to run an ante-mortem analysis on a specific module or directory to generate fictional post-mortems and hardening TODOs.
