agentleFS
Badhon-00Badhon-00/MDCP/skills/master-developer/references/CLAUDE.md

MDCP / references

CLAUDE.md2 starsChanged 5 days agoMITView on GitHub

Project rules for Badhon-00 (Melissa projects: Next.js, Node/Express, TypeScript, Prisma/PostgreSQL, Redis).

  • Reads credentials

The file

We split this CLAUDE.md into the rules it gives an agent

11 rules. About 0 are common, 6 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Add new features in new files/modules.

    line 6
  2. Distinctive:

    Read the code you touch and its callers before editing.

    line 7
  3. Distinctive:

    Ask one short question only when a wrong guess is expensive or irreversible.

    line 8
  4. Distinctive:

    Never report something as built, tested or secure unless it was actually run.

    line 9
  5. Very specific:

    No secrets in code; config goes in .env with a .env.example.

    line 14
  6. Very specific:

    Run npm audit and fix high/critical findings.

    line 22
  7. Very specific:

    Put all DB and process.env secret access in one server-only Data Access Layer (import 'server-only').

    line 30
  8. Very specific:

    Treat every Server Action and route.ts as a public HTTP endpoint: validate input and re-check auth inside each one.

    line 31
  9. Very specific:

    Only NEXT_PUBLIC_ env vars reach the browser.

    line 33
  10. Distinctive:

    Use the existing design tokens and components.

    line 38
  11. Distinctive:

    Read related files in one batch.

    line 46

≈ “Common” is an estimate: at least 1 in 5 of the 27,641 CLAUDE.md files we read have a rule of the same kind, not the same words. “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
CLAUDE.md

CLAUDE.md

Project rules for Badhon-00 (Melissa projects: Next.js, Node/Express, TypeScript, Prisma/PostgreSQL, Redis).

Working rules

  • Add new features in new files/modules. Edit existing working files only when the feature needs it, and keep the diff small.
  • Read the code you touch and its callers before editing. Follow the structure the project already has; do not add files the owner did not ask for.
  • Ask one short question only when a wrong guess is expensive or irreversible. Otherwise pick the sensible default, say which one, and continue.
  • Never report something as built, tested or secure unless it was actually run. State what was not checked.

Done means

  • npm run build and npx tsc --noEmit pass.
  • Tests for new non-trivial logic exist and pass.
  • No secrets in code; config goes in .env with a .env.example.

Security baseline

  • Validate every input at the boundary (Zod). Use Prisma/parameterized queries only.
  • Auth and role checks on every protected route. Rate-limit public and auth endpoints.
  • Passwords: Argon2id. Sessions/JWT: httpOnly cookies, rotating refresh tokens.
  • Secure headers (CSP, HSTS, X-Content-Type-Options, frame protection) set in config.
  • Sanitize HTML before rendering. Allowlist URLs for any server-side fetch (SSRF).
  • Run npm audit and fix high/critical findings.

Checklist: OWASP Top 10:2025 (owasp.org/Top10/2025)

A01 Broken Access Control, A02 Security Misconfiguration, A03 Software Supply Chain Failures, A04 Cryptographic Failures, A05 Injection, A06 Insecure Design, A07 Authentication Failures, A08 Software or Data Integrity Failures, A09 Logging and Alerting Failures, A10 Mishandling of Exceptional Conditions.

  • A03: pin and review dependencies, commit the lockfile, no unvetted install scripts.
  • A10: every error path fails closed (deny, not allow) and never leaks stack traces or internals to the client.

Next.js (nextjs.org/docs/app/guides/data-security)

  • Put all DB and process.env secret access in one server-only Data Access Layer (import 'server-only'). It authorizes, then returns minimal DTOs, never raw rows.
  • Treat every Server Action and route.ts as a public HTTP endpoint: validate input and re-check auth inside each one.
  • Client Component props take only the fields they render. Route params and searchParams are user input.
  • Only NEXT_PUBLIC_ env vars reach the browser; never put secrets there.
  • Behind a reverse proxy, set serverActions.allowedOrigins. Mutations never happen during render.
  • Audit middleware.ts and route.ts first.

UI

  • Use the existing design tokens and components. Cover loading, empty, error and success states.
  • Respect prefers-reduced-motion; keep animation at 60fps (transform/opacity, instancing for 3D).

More rules

Topic files live in claude-rules/ (index: claude-rules/master-manifest.md). Open the one that matches the task; do not load all of them.

Token use

  • Show diffs or changed snippets, not whole files, unless the file is new.
  • Read related files in one batch. Do not re-read unchanged files.

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.