
Bug Receipt
Closes bug and incident work with a standard receipt marking it VERIFIED, PARTIAL, or BLOCKED.
- Repository
- github/awesome-copilot
- Stars
- 40,703
- Licence
- MIT
- Files
- 5 files
What it does
After an agent diagnoses, fixes, or recovers from a software defect or incident, this skill makes it end with a fixed-format closeout report. The report records the observed problem, the failing check, the root cause, the change made, the proof that it works, any gaps, and whether the evidence came from this run or was supplied. The agent is also told to keep sensitive data like tokens and personal details out of that evidence, and to avoid calling a fix proven unless the actual user-visible behavior was checked.
Great for
- Developers who want a consistent, auditable record of how a bug was closed
- Teams using AI agents for incident follow-up who need to see what was actually executed versus assumed
- Engineers who want machine-readable closeout JSON validated in CI
Skip it if
- Quick questions or non-bug tasks, since the skill forces a full receipt format on every closeout (the skill says to return it even when the user asks for brevity)
- Situations where you need the agent to freely summarise without a rigid template
Try asking
- “Fix the 10% discount rounding bug and close it with a bug receipt.”
- “Recover from this production incident and give me the BUG RECEIPT with the evidence source marked.”
- “Produce a machine-readable bug receipt JSON for this fix and validate it with the bundled script.”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 5 original files, the MIT licence and an attribution note.
bug-receipt.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip bug-receipt.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip bug-receipt.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Give the agent the failing command or reproduction steps up front so the Baseline row can be filled with a real result.
- Check the Source row: 'supplied' means the agent did not run the proof itself.
- Review the validation script before running it; it reads files you pass to it, and the bundled excerpt shows no network calls.
What's inside
- assets/receipt.template.json
- references/receipt-contract.md
- references/receipt.schema.json
- scripts/validate-receipt.mjs
- SKILL.md
Redistributed unmodified from github/awesome-copilot under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.