AGENT SKILL

de-slop

Use when cleaning a diff before opening a PR, or when asked to simplify, de-slop, or run the full quality pass on current changes.

License 许可
MIT
最近更新
2026-11-15
Stars
★ 0
Forks
1

01 前言 Front Matter

name
de-slop
triggers
[ "user", "model" ]
description
Use when cleaning a diff before opening a PR, or when asked to simplify, de-slop, or run the full quality pass on current changes.

03 SKILL.md 正文

De-Slop

One pass that sequences the bundle's simplification skills over the current diff or branch before a PR. Each invoked skill keeps its own rules; this skill only fixes the order and the stop conditions.

Scope

The current diff against the base branch (default main), including the working tree, or the scope the caller names.

Codebase mode: when the caller names the whole codebase (or a subtree) instead of a diff - "untangle this codebase", "clean up the accumulated slop" - hunt what long agent-assisted sessions pile up: duplicated logic copied instead of shared, wrapper layers that only forward calls, temporary workarounds left in past their fix, dead paths kept "just in case". Pair with big-file-refactor when a file has crossed ~1k lines; run the same run order below on the found set, not on a diff.

Run order

Apply each skill to the scoped diff, in this order:

  1. lean-code: minimal-code sweep. Delete unneeded code, options, and abstractions; prefer the smallest complete change.
  2. no-comments: comment-smell pass. Hunt comments, apply the accept/reject rubric, land smallest root-cause fixes, encode claimed constraints as type, test, or lint where approved.
  3. unslop: prose pass on changed docs, comments that survive, commit text, and PR body drafts.
  4. thermo-nuclear-code-quality-review: structural pass over the whole diff. Use thermos instead when the dual review is wanted.
  5. Conditional language passes, only when the diff touches them:
    • typescript-best-practices for .ts changes.
    • vercel-react-best-practices for .tsx or React changes.

Re-read the diff after each step; later passes see the code earlier passes left behind. Freeze the scope at step 1: new findings outside it are reported open, not fixed.

Multi-pass guidance

The skills are non-deterministic: a second run of the whole sequence catches findings the first run missed. Run the pass, then run it once more on the resulting diff. A third run is worthwhile only while the previous run still produced accepted changes.

Stop conditions

Stop when either holds:

  • A full pass produces no new accepted findings or edits.
  • The diff is frozen for review (caller says stop, or PR is open).

Report

After the final pass, report per skill: findings accepted and rejected, edits made, encodings applied or offered, open work, and which pass (number) produced each change.

04 同仓库更多 Skills

Leostruka
Leostruka/devin-bundle

montador-de-ideias

Use when assembling, expanding, or elevating a raw project idea or initial prompt into an ultra-detailed, first-principles dossier and compiled master prompt with journalistic OSINT market research, Tech Lead and Staff Architect justifications, atomic and modular hardening/optimization, and high-impact contact surfaces.

★ 0
Leostruka
Leostruka/devin-bundle

git-github-adaptive-engineering

Use when any task touches git history, GitHub issues/PRs/releases/CI, repo governance, or branch and merge decisions. Runs the Fase Zero situational-awareness battery, picks one of five delivery flows via the adaptive decision engine, and enforces commit/PR information-density standards. Orchestrates the gh, git-workflows, github-actions, pr, and debug/finish skills.

★ 0
Leostruka
Leostruka/devin-bundle

adaptive-agenda-calendar-routine-engine

Use when planning or executing work across hours, days, weeks, or sprints. Turns the calendar into a strict time-blocked agenda (HH:MM-HH:MM blocks carrying a routine profile, expected inputs, and a handoff output artifact), runs the 5-stage cognitive flow (Radar and Hunt -> Triage plus Steve Jobs conception -> workbench manifest -> >=32-cycle loop -> Close plus Open-Minded Scrum), and unfolds macro calendar milestones into micro agenda blocks through extensions/gcal/wrapper.py.

★ 0
Leostruka
Leostruka/devin-bundle

demo

Use when demonstrating a skill in a fixture project.

★ 0
Leostruka
Leostruka/devin-bundle

duplicate-skill

Use when testing duplicate skill name detection.

★ 0
Leostruka
Leostruka/devin-bundle

duplicate-skill

Use when testing duplicate skill name detection.

★ 0