minimal-run-and-audit
installsforks
How to install minimal-run-and-audit?
lllllllama/ai-paper-reproduction-skill/minimal-run-and-audit
$npx -y skills add lllllllama/ai-paper-reproduction-skill --skill minimal-run-and-auditminimal-run-and-audit is an agent skill published from lllllllama/ai-paper-reproduction-skill, installable through the skills CLI.
SKILL.md
| 1 | # minimal-run-and-audit |
| 2 | |
| 3 | Use this as the Rigor Run skill. The installed slug remains |
| 4 | `minimal-run-and-audit` for compatibility. |
| 5 | |
| 6 | Use the shared operating principles in |
| 7 | `../../references/agent-operating-principles.md`; this skill should make run |
| 8 | evidence auditable without turning every command into a rigid protocol. |
| 9 | |
| 10 | ## When to apply |
| 11 | |
| 12 | - After a reproduction target and setup plan exist. |
| 13 | - When the main skill needs execution evidence and normalized outputs. |
| 14 | - When a smoke test, documented inference run, documented evaluation run, or other short non-training verification is appropriate. |
| 15 | - When the user already knows what command should be attempted and wants execution plus reporting only. |
| 16 | |
| 17 | ## When not to apply |
| 18 | |
| 19 | - During initial repo scanning. |
| 20 | - When environment or assets are still undefined enough to make execution meaningless. |
| 21 | - When the task is a literature lookup rather than repository execution. |
| 22 | - When the user is still deciding which reproduction target should count as the main run. |
| 23 | |
| 24 | ## Clear boundaries |
| 25 | |
| 26 | - This skill owns normalized reporting for an attempted command. |
| 27 | - It may receive execution evidence from the main skill or a thin helper. |
| 28 | - It does not choose the overall target on its own. |
| 29 | - It does not perform broad paper analysis. |
| 30 | - It does not own training startup, resume, or long-running training state. |
| 31 | - It should not normalize risky code edits into acceptable practice. |
| 32 | - It must not hide changes that alter evaluation, preprocessing, checkpoints, |
| 33 | metrics, or other scientific meaning. |
| 34 | |
| 35 | ## Input expectations |
| 36 | |
| 37 | - selected reproduction goal |
| 38 | - runnable commands or smoke commands |
| 39 | - environment and asset assumptions |
| 40 | - optional patch metadata |
| 41 | |
| 42 | ## Output expectations |
| 43 | |
| 44 | - execution result summary |
| 45 | - standardized `repro_outputs/` files |
| 46 | - `SCIENTIFIC_CHANGELOG.md` for changed scientific meaning and evidence status |
| 47 | - `COMPARABILITY_REPORT.md` for README/paper/baseline comparability |
| 48 | - clear distinction between verified, partial, and blocked states |
| 49 | - `PATCHES.md` when repo files changed |
| 50 | |
| 51 | ## Notes |
| 52 | |
| 53 | Use `references/reporting-policy.md`, `../../references/research-rigor-principles.md`, `scripts/run_command.py`, and `scripts/write_outputs.py`. |
People who liked this also viewed
6 picksType
- find-skillsHelps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skSkillsJul 2026
- grill-meA relentless interview to sharpen a plan or design.SkillsJul 2026
- vercel-react-best-practicesVercel's official collection of agent skillsSkillsJul 2026
- improve-codebase-architectureScan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.SkillsJul 2026
- setup-matt-pocock-skillsConfigure this repo for the engineering skills — set up its issue tracker, triage label vocabulary, and domain doc layout. Run once before first use of the other engineering skills.SkillsJul 2026
- triageMove issues and external PRs through a state machine of triage roles — categorise, verify, grill if needed, and write agent-ready briefs.SkillsJul 2026
