| name | papercuts |
| description | A tiny "papercuts" CLI tool that agents can use to complain about bullshit they encounter during work, like dead-end tool calls, broken links, or other frustrations. Use it proactively when small, non-blocking friction reveals something in a repo or workflow that should be sanded down later. |
Log papercuts
Record small friction immediately without interrupting the task.
- Write one or two sentences: what you were doing, what got in the way, and, when useful, a likely cause or fix.
- Identify both your agent runtime and exact model ID. Never shorten, generalize, or guess the model ID.
- Run
papercuts --agent <your-agent-name> --model <exact-model-id> "<message>" from the directory where the friction occurred. Omit --model only when the runtime exposes the exact ID for automatic detection. When the friction fits an obvious category (for example flaky-command, broken-link, docs, stale-cache, misleading-error), add it with --tag <category> so recurring friction is countable.
- Continue the original task. Mention the papercut to the user only when it materially affects the result.
Use this for missed or dead-end tool calls, broken links, confusing or undocumented setup, flaky commands, stale caches, misleading errors, and non-obvious gotchas.
Do not use PAPERCUTS.md as a substitute for a real bug report, tracked work, or the final account of what was accomplished. Never include secrets, tokens, private user data, or raw credentials.
The CLI writes to PAPERCUTS.md at the enclosing Git root and records the timestamp, invocation directory, agent runtime, and exact model ID. If papercuts is unavailable, continue the task and briefly report that the logging tool is not installed; do not create a different log format by hand.