What problem does it solve? Long-running or unattended agent work leaves reviewers unable to reconstruct what was decided, why, and on what evidence without rereading the entire transcript. This Skill keeps a single canonical decision trail so a human can audit the run after the fact. ## Core Features & Use Cases - Structured TSV decision log: One row per decision with timestamp, phase, decision, why, evidence, and result columns, rendered as a sortable table on GitHub or in a terminal. - Safe logging helper: The scripts/log.sh script stamps timestamps, writes the header on first use, strips stray tabs and newlines, and escapes spreadsheet formula characters in generated cells. - End-of-run audit and cross-model review: The log is checked against the run transcript for truthfulness, and a subagent on a different model family flags weak evidence, skipped verification, and risky choices in an Attention section. - Use Case: During a multi-week migration, log each fork chosen, unit completed, and revert triggered to decisions.tsv, then commit the log so the PR reviewer can follow the evidence pointers row by row. ## Quick Start Ask the agent to start a show-me-your-work decision log for this task and record each major decision with its evidence as the run progresses.
