agent-harness-design
Designing agent harnesses and tool systems — risk taxonomy for tools, permission decisions, draft/commit pattern, structured tool…
Remove AI-generated code noise from the current diff while preserving behavior. Use for deslop, AI slop cleanup, unnecessary comments, defensive checks, any casts, or needless nesting after an agent-generated change. Do not use as permission for a broad rewrite or when a real
$ npx -y skills add AnastasiyaW/codex-claude-code-config --skill deslop --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/deslopContext preview
The summary Claude sees to decide when to auto-load this skill.
Remove AI-generated code noise from the current diff while preserving behavior. Use for deslop, AI slop cleanup, unnecessary comments, defensive checks, any casts, or needless nesting after an agent-generated change. Do not use as permission for a broad rewrite or when a real
name: deslop description: Remove AI-generated code noise from the current diff while preserving behavior. Use for deslop, AI slop cleanup, unnecessary comments, defensive checks, any casts, or needless nesting after an agent-generated change. Do not use as permission for a broad rewrite or when a real bug has not been isolated.
Review only the branch diff against its intended base and remove noise that does not belong in the local code style. Keep the behavior and public contract unchanged unless a clear, separately verified bug is fixed.
module;
established RAII/contract boundary.
1. Inspect the base, diff, local style, tests, and ownership boundaries. 2. Classify each candidate as noise, a clear bug, or an intentional contract. 3. Remove only confirmed noise in a focused edit. 4. Run the narrow relevant checks and inspect the final diff. 5. If the structure needs a real redesign, stop deslop and use `refactoring-safely`, `architecture-quality`, or `thermo-nuclear-code-quality-review`.
Do not delete comments that explain a non-obvious invariant, security boundary, ABI constraint, workaround with an owner, or externally required behavior.
carry domain meaning.
cleanup into a regression.
ABI or lifetime boundary; inspect callers before removing it.
| Symptom | Likely cause | Action | |---|---|---| | Cleanup changes a test result | Candidate was behavior, not noise | Revert that candidate and isolate the real contract | | Diff is too broad | Tool ran over the whole tree | Restrict review to the branch diff and restore unrelated files | | Comment seems redundant but explains a constraint | Context is outside the file | Read the owning docs/tests before changing it | | Code remains structurally tangled | Deslop is the wrong scope | Escalate to a planned refactor with characterization tests |
Adapted from Cursor Team Kit's MIT-licensed `deslop` workflow: https://github.com/cursor/plugins/tree/main/cursor-team-kit/skills/deslop
A practical configuration kit for Claude Code, Codex, and other coding agents. It contains architectural principles, enforcement hooks, skills, drop-in rules, starter templates, and dynamic-workflow commands.
Designing agent harnesses and tool systems — risk taxonomy for tools, permission decisions, draft/commit pattern, structured tool…
Use when designing, auditing, refactoring, or explaining an agentic harness for any domain, especially when work must continue…
Практическая инженерия диффузионных моделей: архитектуры, обучение, инференс, оптимизация памяти. Использовать при любых задачах…
Expert prompt engineering for FLUX.2 [klein] image generation and editing model. Use this skill whenever the user wants to create…
Plan or review LoRA and edit-training work specifically for FLUX.2 Klein or Qwen-Image-Edit, including paired datasets,…
Forensic image-to-prompt compiler for image generation models. Use this skill whenever the user wants to: convert/describe an…