clean-code
Engineering guardrails for writing, editing or reviewing source code, so an edit leaves no residue, keeps every type and reads like the code around it.
Skill by osvfelices
Chalk Labs
Need help deploying clean-code?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/osvfelices-clean-code
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: writing, productivity
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/clean-code
What's inside
- Skill name
clean-code- Description
- Engineering guardrails for writing, editing or reviewing source code, so an edit leaves no residue, keeps every type and reads like the code around it. Use on any coding task that creates or changes source code.
- License
- MIT
- Instructions
- 26 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# clean-code After each edit, and after each shell command in a Git worktree, a hook reads the lines you just wrote in TypeScript, TSX, JavaScript, JSX or Python, and in other languages on a best-effort basis. It lists what to fix by line, says `not checked` when it cannot read a change, and says nothing otherwise. Fix what it lists, in the file you are editing; leave older code alone unless asked. `not checked` is not clean: read those lines yourself. A shell command is checked on the source changes visible when its tool reports completion, and its safeguards only after it ran: if the hook says it weakened one, undo it. Writes that occur after the tool completion event, including detached or background descendants, are outside that observation boundary. Tracked and Git-visible untracked files are covered; Git-ignored paths are outside shell coverage unless they are already tracked, and `.clean-code.json` exclusions stay unchecked as before. Writes outside the worktree and other tools' writes are not seen. What follows covers what a hook cannot see. ## Comments Keep a comment that gives intent, a consequence, a constraint the code cannot show, or a public contract. Delete one that narrates the edit or repeats the code; rewrite a useful one that reads badly. Plain sentences: the hook flags decorative glyphs, all-caps emphasis, task or phase labels, and file headers past four lines. ## Names Reveal intent so the reader can skip the body. One word per concept, in the project's own vocabulary. Nouns for types, verbs for functions, predicates for booleans. No encodings, no `Manager`/`Helper`/`Utils` dumps. ## Functions One job at one level of abstraction, callers above callees. Keep required parameters few; group related ones into an object. A boolean that picks between two behaviors usually wants two functions; a boolean that is data is fine. Change state or answer a question, not both. No hidden side effects or temporal coupling. Extract only when the new name reads better than the body. ## Types and errors Never remove or loosen an existing type. Validate untrusted input once, at the boundary: parameterized queries, allowlists. Errors carry context. Return an empty value, a result type or an exception where `null` would hide the case. Handle edges: empty, zero, one, last, max, malformed. ## Scope Implement what was asked, nothing speculative: no single-use abstractions, no unrequested files, no config "for later". Extract duplicated knowledge, not code that merely looks alike. Follow the project's formatter, linter and test style. Do not reformat lines you are not changing. ## Tests New behavior ships with a test named after the behavior. Fast, independent, repeatable, one concept each. ## Judgment A linear 30-line function can beat six fragments. Idioms differ by language: Go's explicit errors are right in Go. When asked to cut a corner, state the cost in one line, then do it.
From osvfelices/clean-code, MIT.
More from osvfelices/clean-code
- 0.546score
clean-options
osvfelices
Propose 2-3 design approaches with trade-offs for a task and wait for the user's choice.
SkillBFreshness: B★ 2Claude Code +9 - 0.521score
clean-sweep
osvfelices
Run the clean-code checker on the working tree or a given path and fix what it reports.
SkillBFreshness: B★ 2Claude Code +9
Signals
- Stars
- 2
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Sep 25, 2026 (1 month ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/osvfelices/clean-code
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from osvfelices
- 0.546score
clean-options
osvfelices
Propose 2-3 design approaches with trade-offs for a task and wait for the user's choice.
SkillBFreshness: B★ 2Claude Code +9 - 0.521score
clean-sweep
osvfelices
Run the clean-code checker on the working tree or a given path and fix what it reports.
SkillBFreshness: B★ 2Claude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get clean-code running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
clean-code skill. Agent Mart. https://agentmart.dev/skills/osvfelices-clean-code (last updated 2026-10-24)