Skip to content
Back to skills

Intuitive Squash

ASecurity

Safely rewrite noisy local history into reviewable commits, with aggressive or moderate plan choices, when the user explicitly asks to squash.

  • 48 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 11, 2026
ai-agentssecurity

Security analysis

A100/100

Pro scans all 6 files and shows the line behind each finding

Scanned October 11, 2026

npx -y skills add MiaoDX/intuitive-flow --skill intuitive-squash --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Intuitive Squash?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Intuitive Squash
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/miaodx-intuitive-squash/badge)](https://www.skillsdirectory.com/skills/miaodx-intuitive-squash)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: intuitive-squash
description: Safely rewrite noisy local history into reviewable commits, with aggressive or moderate plan choices, when the user explicitly asks to squash.
disable-model-invocation: true
---

# Intuitive Squash

Turn noisy local history into reviewable commits without changing the final tree.
Read [planning](references/planning.md) to resolve the base, run the bundled
read-only preflight, identify preserved commits, and propose the commit map.
Do not mutate branches or the worktree during planning.

## Modes

Offer both modes by default, unless the user requests one exact strategy:

| Mode | Commit shape |
| --- | --- |
| **Aggressive** | The fewest coherent reviewable commits, while keeping preserved commits and unrelated change surfaces separate. |
| **Moderate** | Semantic review and rollback boundaries with phase and fixup churn removed; prefer this for large or high-risk branches. |

State the recommended mode and why, then use the mode the user approves for the
commit map. For a small stack, the plans may differ only slightly; do not invent
splits to make them look different.

Keep marked, security/hotfix, configured-path, and other-author commits standalone
as the preflight requires. Disclose publication, merge, signature, and tag risks;
record the original-to-final mapping, planned HEAD, backup, and stash strategy.
Never rewrite until the user approves the exact plan and applicable risk gates.

Once approved, read [rewrite and restore](references/rewrite-and-restore.md).
Recheck planned state, back up before rewriting, verify final tree equality and
preserved commit mappings, then restore dirty work. Keep the backup; retain the
stash on verification failure or restoration conflict. Report proof and recovery
state. Push only when separately authorized, with an explicit expected-old-OID
lease if rewriting published history.

Files in this skill

  • SKILL.md1.8 KB
  • agents/openai.yaml317 B
  • references/planning.md5.6 KB
  • references/rewrite-and-restore.md2.4 KB
  • scripts/preflight.test.ts6.6 KB
  • scripts/preflight.ts14 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…