agentleFS
augments-labsaugments-labs/sdlc-skills/skills/design/coding-standards/SKILL.md

coding-standards

Skill1 starsChanged 3 months agoMITView on GitHub

Settles the conventions agents and humans both follow in a project: domain vocabulary, naming, patterns to reach for, and things to never do. Use when a project needs its conventions settled, once per project or when they have drifted, or when the user asks for a style guide or house rules, says the codebase is inconsistent, or asks how something should be named. Skip when clear standards already exist and are followed, and skip proving the code correct.

120 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use everything-claude-code, everything-claude-code-conventions and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • InfopibeSkill

    everything-claude-code

    Also in 9 projects that use this

  • mryeastcodeSkill

    everything-claude-code-conventions

    Development conventions and patterns for everything-claude-code. JavaScript project with conventional commits.

    Also in 8 projects that use this

  • niloykumarbarmanSkill

    agent-sort

    Build an evidence-backed ECC install plan for a specific repo by sorting skills, commands, rules, hooks, and extras into DAILY vs LIBRARY buckets using parallel repo-aware review passes. Use when ECC should be trimmed to what a project actually needs instead of loading the full bundle.

    Also in 8 projects that use this

  • tpavanipradeepSkill

    nextjs-turbopack

    Next.js 16+ and Turbopack — incremental bundling, FS caching, dev speed, and when to use Turbopack vs webpack.

    Also in 15 projects that use this

  • CodeWithBehnamSkill

    dmux-workflows

    Multi-agent orchestration using dmux (tmux pane manager for AI agents). Patterns for parallel agent workflows across Claude Code, Codex, OpenCode, and other harnesses. Use when running multiple agent sessions in parallel or coordinating multi-agent development workflows.

    Also in 17 projects that use this

  • InfopibeSkill

    investor-outreach

    Draft cold emails, warm intro blurbs, follow-ups, update emails, and investor communications for fundraising. Use when the user wants outreach to angels, VCs, strategic investors, or accelerators and needs concise, personalized, investor-facing messaging.

    Also in 22 projects that use this

  • +1 more from Infopibe

The file

We split this skill into the rules it gives an agent

8 rules. 0 are common, 5 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 104 edited versions we read keep these.

  • 5

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Skip: if the project already has clear standards that are actually followed.

    line 11
  2. Very specific:

    If an approved data-model or domain contract exists, take its concepts and terms.

    line 17
  3. Distinctive:

    If there is no contract, derive the terms from the domain evidence you can see.

    line 22
  4. Distinctive:

    Write the hard nevers, the things never to do in this project.

    line 29
  5. Distinctive:

    Never fabricate a path.

    line 41
  6. Very specific:

    Write the immutable section (the standard itself: once issued, a change to it is a successor version, never an edit) to …

    line 50
  7. Very specific:

    Ask this and end the turn.

    line 69
  8. Distinctive:

    If the answer is option 1, claim adoption only after enforcement and the exemplar gate (the exemplar's checks and rubric from Step 2.3) …

    line 70

“Common” is measured: kept by at least 6 in 10 of the 104 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Coding Standards

When to use

  • Starting a project, or when conventions have drifted and the code reads like several authors.
  • Skip if the project already has clear standards that are actually followed.

Step 1: Settle the vocabulary and the patterns

Open assets/standards-template.md before starting; each step fills its section. Read references/standards-example.md when the first draft is written and needs comparing against a filled example.

  1. If an approved data-model or domain contract exists, take its concepts and terms. Decide only their representation (how each term appears in code), casing, abbreviations, and the enforcement that catches drift from them. Never decide their meaning, and never rename one. If there is a dispute about what a concept is, invoke data-model.
  2. If there is no contract, derive the terms from the domain evidence you can see.
  3. Ban a generic label wherever a domain term exists. Keep the glossary current.
  4. Name the one way this project does each of these: error handling, validation, async, dependency injection, and testing seams. Give one short example of each.
  5. Write the hard nevers, the things never to do in this project.

Step 2: Assign enforcement

  1. Assign enforcement to each rule. If the rule can be detected statically, assign a formatter, linter, type or static check, or repository check that fails. If the rule is about vocabulary, architecture, or judgement, assign a named review rubric and a named owner.
  2. The warning budget is zero. Never leave a permissive default.
  3. Point to the exemplar (the file that does all of the above right). Run its checks and rubric now. If the project is greenfield and has no exemplar, name the first artifact and its owner; adoption stays pending. Never fabricate a path.
  4. Name which project instruction wins when instructions conflict, and who approves an exception. Each exception needs a scope, a reason, an expiry, and a compensating check.
  5. Name how often, or on what trigger, the standard is reviewed for drift, and who updates it.

Step 3: Write and present

  1. Write the immutable section (the standard itself: once issued, a change to it is a successor version, never an edit) to .sdlc-skills/designs/{{YYYY-MM-DD}}-{{topic}}.md or the user-set path, preserving approved sections around it.
  2. Present:

    Coding standards {{path}} — version {{identity (per template)}}
    Vocabulary: {{n}} terms  Patterns: {{n}}  Nevers: {{n}}
    Enforcement: {{n}} automated, {{n}} review  Exemplar gate: {{result}}
    
    1. Approve
    2. Request changes
    3. Reject
    4. Cancel
    
    Recommendation: {{option}} — {{one sentence}}.

    Ask this and end the turn; clarifying-intent owns what closes it.

  3. If the answer is option 1, claim adoption only after enforcement and the exemplar gate (the exemplar's checks and rubric from Step 2.3) have run. Record in force / suspended / superseded outside the section, in the external adoption ledger the template names (its record of adoption), with fresh evidence.

Gotchas

  • An exception granted with no expiry never comes back for review. The carve-out becomes a second, permanent convention living alongside the one it was supposed to be an exception to.
  • Enforcement usually checks new or changed code, not a one-time sweep of everything already there, so banning a generic label going forward can leave every existing occurrence exactly as inconsistent as before.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.