# Principle Explicit Intent

> Requires stated intent for irreversible actions. Apply before merge, force-push, close, delete, or publish operations.

- Skill: `bostonaholic/principle-explicit-intent` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add bostonaholic/principle-explicit-intent`
- Raw SKILL.md: https://api.skillmd.com/api/skills/bostonaholic/principle-explicit-intent/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: bostonaholic (https://skillmd.com/u/bostonaholic)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/bostonaholic/principle-explicit-intent

---


# Explicit Intent

Perform merges, pushes over published history, public closes, and deletions only on the user's stated intent, never inferred from state.

- Do not infer ship intent from green, rebase intent from behind, or abandon intent from stale.
- Require one yes per irreversible mutation; a named set authorizes that set, never unnamed items or adjacent change classes.
- Treat authorization as permission to finish the verified act, and nothing beyond it.
- Use granted authorization without re-asking; after gates pass, finish. Bound and report any confirmation churn.
- Put `"Invoke ONLY on … never infer …"` in descriptions of skills whose invocation authorizes side effects; set `disable-model-invocation` for stricter host-enforced entry guards.
- Put the guard at each in-run approval instead when invocation alone authorizes no mutation.