unslop
Remove unnecessary abstractions, defensive code, comments, or prose from a requested diff or files. Preserve behavior and repository rules.
- License 许可
- AGPL-3.0
- 最近更新
- 2026-11-15
- Stars
- ★ 635
- Forks
- 91
01 前言 Front Matter
- name
- unslop
- description
- Remove unnecessary abstractions, defensive code, comments, or prose from a requested diff or files. Preserve behavior and repository rules.
02 资源文件 5
- 参考文档references/patterns.md12.9 KB
- 参考文档references/semantic-search.md7.2 KB
- 参考文档references/vector-search.md6.4 KB
- 参考文档UPSTREAM-LICENSE.txt1.1 KB
- 参考文档UPSTREAM-NOTICE.md1.6 KB
03 SKILL.md 正文
Unslop
A pre-commit pass that removes patterns LLMs add reflexively but humans wouldn't write. Apply it to the requested diff or files. The goal isn't to shorten the diff — it's to delete code whose absence would not be missed.
This skill complements (does not replace) AGENTS.md and web/src/components/ui_primitives/STRATEGY.md. Those define the rules. This skill defines the patterns to actively hunt and remove.
Unslop is a quality pass, not a bug hunt. For correctness review of a diff, branch, or PR — including the NodeTool-specific landmines — use code-review; a full pre-merge pass runs both.
How to use
- After making changes, run
git diffand read every added line through the lenses below. - For each "slop" you find, delete or rewrite it. Don't leave a
// removed Xcomment behind. - Complete mandatory post-change verification for code edits. For prose-only edits, check affected links and formatting. Reuse valid results for unchanged files.
- When deleting an abstraction, remove unused types and exports. Preserve behavior coverage by moving useful tests to the surviving public interface.
Pattern reference
Read patterns.md for the languages and frameworks in the diff. Repository rules take precedence over these cleanup heuristics.
Self-review checklist
Use the relevant questions to inspect the requested diff. Verify each concern against the repository rules before changing code.
- [ ] Did I add a comment that restates the code, names the PR, or describes a removed feature?
- [ ] Did I add a
try/catchwhose error path can't actually trigger here? - [ ] Did I write
any,as any, oras unknown asto silence the compiler? - [ ] Did I add
useEffectto compute a value from props/state I already have? - [ ] Did I
useCallback/useMemo/React.memowithout a memoized consumer or measurable cost? - [ ] Did I subscribe to a whole Zustand store (
const s = useFooStore()) or skipuseShallowon a multi-key selector? - [ ] Did I import a raw MUI component into a non-primitive file, or hardcode a color, spacing, radius, font size, or transition?
- [ ] Did I write a
useEffect+fetchinstead ofuseQuery? - [ ] Does any new test assert implementation rather than user-visible behavior?
- [ ] Are there
// TODO,// removed,// added by, or "useful elsewhere" leftovers? - [ ] Could three near-identical lines have been left as-is instead of becoming a helper?
- [ ] Does my prose contain any banned openers or filler ("delve", "robust", "seamlessly", "Here's the thing")?
- [ ] Have I left any required verification incomplete?
Report the concrete removals and verification results. A checklist alone does not establish correctness.
Sources
Patterns synthesized from these community skills, adapted to NodeTool's stack:
- theclaymethod/unslop — humanizing AI-generated prose
- hardikpandya/stop-slop — banned phrases and structural red flags
- jalaalrd/anti-ai-slop-writing — banned-word and banned-pattern catalog
- anthropics/skills/frontend-design — anti-generic-aesthetic principles
- vercel-labs/agent-skills/composition-patterns — composition over boolean-prop proliferation
- vercel-labs/agent-skills/react-best-practices — React/Next performance rules
- awesome-skills/code-review-skill — React 19 / TypeScript review patterns
- NodeTool's own
AGENTS.mdandSTRATEGY.md
04 同仓库更多 Skills
humanizer
Rewrite AI-sounding text so it reads like the writer without changing what it says. Use when editing or reviewing prose for AI tells: not-X-but-Y contrasts, one-line closers, staged openers, forced triads, dashes everywhere, inflated claims, sales language, stock AI words, bold labels, or filler.
nodetool
Use the NodeTool MCP server to build, validate, run, and debug visual AI workflows, and to work with assets, collections, and jobs. Use when the user mentions NodeTool or asks for a workflow graph, a node, or a media generation run on their NodeTool server.
wan-2-6-prompting
Prompt Alibaba's Wan 2.6 video — text-to-video with a global style line plus timing-bracketed shots, and image-to-video that describes only what changes over time. Covers the shot-mode switch per provider, the prompt budget, negative prompts, motion intensity wording, audio input and prompt expansion. Use whenever the model id names a Wan 2.6 video route (alibaba/wan-2.6/text-to-video, alibaba/wan-2.6/image-to-video, wan-video/wan-2.6-t2v, wan-video/wan-2.6-i2v, wan-video/wan2.6-i2v-flash) on generate_video, animate_image or a TextToVideo node. Not for Wan 2.5 or 2.7, or the Wan 2.6 image routes.
video-workflow
Build a reusable NodeTool video workflow for batches or repeated inputs. For a single finished piece, use a storyboard skill.
video-clone
Analyze a reference video and recreate its shot structure with a new product or cast in NodeTool.
video-audio-continuity
Keep sound continuous across a multi-scene piece cut from generated video — why one clip per scene hard-cuts the audio at every boundary, when to write all the scenes into a single generation instead, the sound-design vocabulary that survives a provider's copyright filter, and how to check the length, the aspect ratio and the mix of what came back. Use when a piece has more than one scene and the video model writes its own audio (Seedance 2, Veo 3, MiniMax H3, Kling 2.6 and later), or when a cut's sound drops out at the shot changes.