Agent skill · shipshitdev
show-me-your-work
Keep a reviewable decision trail for long-running or unattended work. A TSV log with one row per decision (what, why, evidence, result). Local by default. Commit it when a reviewer needs the trail to trust the result. Use for show-me-your-work, autonomous or multi-phase runs, or work a human reviews after stepping away.
What it needs
About 2k tokens when loaded.
What this skill does
Show me your work For work a human reviews after the fact, a decision trail reconstructs what was decided, why, and on what evidence. Authorized Scope Apply this engine only within the user's requested task and existing explicit authorization. Loading or delegating to it grants no additional authority. Preserve report-only restrictions and the caller's target, host, provider, and cost limits. Existing approval satisfies a gate only for the same actions and scope; obtain approval before expanding them. Forward these limits to delegates. Contract Inputs: A long, autonomous, or multi-phase run that needs an audit trail Outputs: One canonical TSV log, plus an Attention section from a different-tier review of the trail Creates/Modifies: decisions.tsv in the work dir, or .tmp/audit/<task-slug>.tsv when several efforts run at once Commit the log only when a reviewer needs it External Side Effects: None unless the caller commits the trail Confirmation Required: Before committing the log to the repo Delegates To: None. Other skills route their trail here. Format Copy references/decision-log-template.tsv to start. Columns: ts, phase, decision, why, evidence, result. Cells stay single-line. Evidence is a pointer, not prose. Use scripts/log.sh <logfile> <phase> <decision> <why> <evidence> <result> so rows stay well-formed. Log decision points and checkpoints, not every action. Append-only. A wrong call gets a new row. Write each entry the way you'd tell a teammate. Apply references/prose-slop.md from the selected deslop skill directory to log text. Audit the log Before handing back, walk the log against what actually happened in this run's transcript (the path the harness names). Cut invented rows. Add missing forks. Drop padding. Cross-tier review Spawn a subagent on a different capability tier or family from the one that did the work. It reads the trail and the transcript, then flags weak evidence, skipped verification, and risky choices. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills shipshitdev/show-me-your-work