# Run Deterministic Preflight

> Run the configured seven-layer preflight and mutation proof before a change is presented for review.

- Skill: `homenshum/run-deterministic-preflight` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds@latest add homenshum/run-deterministic-preflight`
- Raw SKILL.md: https://api.skillmd.com/api/skills/homenshum/run-deterministic-preflight/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: HomenShum (https://skillmd.com/u/homenshum)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/homenshum/run-deterministic-preflight

---


# Run deterministic preflight

Read [the execution contract](references/contract.json) before starting. Validate the final receipt with its parameterized `verification.argv`, replacing `{python}`, `{skill_dir}`, and `{receipt}` with concrete local values.

Read the project preflight configuration, then run the checked-in gate. Report
the interpreter, config path, each layer result, aggregate exit code, and the
mutation-proof result.

- Exit `0` is `PASS` only when every applicable layer passes.
- Exit `1` is an observed `FAIL` and must name the failing layer.
- Exit `2` is `NO_GATE`; missing configuration or proof never passes.

Never weaken a check or update an expected value merely to turn red green.
After a repair, rerun the exact mutation that failed and the surrounding test
suite. Preserve bounded raw output as the receipt.