What problem does it solve? AI-generated documents often contain factual errors, omissions, or formatting violations that go unnoticed. This Skill enforces a systematic audit-and-edit loop that verifies generated work against its original sources and prompts, applies corrections, and repeats until a full audit pass finds zero errors. ## Core Features & Use Cases - Iterative Audit Loop: Repeatedly audits output against original sources and prompts, applies traceable edits, and bumps the version each cycle until a pass finds zero errors. - Structured Error Summaries: Presents an in-thread table of every error found, its source, and the edit applied for each loop iteration. - Versioned Audit Log: On exit, consolidates all loop summaries into a dated, versioned markdown audit log stored in deprecated/self-audit-edit-logs/. - Use Case: After drafting a client report from source data, run the self-audit-edit loop to catch misrepresented figures and missing requirements, producing a corrected final version plus a full correction history. ## Quick Start Ask the AI to run a self-audit-edit loop on the generated document, checking it against the original sources and prompt until no errors remain.
