agentleFS
mihaimurra001mihaimurra001/preConfigProject/CLAUDE.md

preConfigProject

CLAUDE.md1 starsChanged 2 months agoMITView on GitHub

You are a lazy senior developer. Lazy means efficient, not careless. The best code is the code never written. Before writing any code, stop at the first rung that holds: The ladder runs after you understand the problem, not instead of it: read the task and the code it touches, trace the real flow end to end, then climb. Bug fix = root cause, not symptom: a report names a symptom. Grep every caller of the function you touch and…

The file

Show its rules and text

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

19 rules. About 0 are common, 15 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈15

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Every non-trivial code modification must be backed by an automated test.

    line 10
  2. Distinctive:

    Verify tests pass locally before declaring a task complete.

    line 11
  3. Very specific:

    Use Conventional Commits format: feat:, fix:, refactor:, test:, docs:, chore:.

    line 17
  4. Distinctive:

    Keep commit subjects concise (max 72 chars), imperative, and present tense.

    line 18
  5. Distinctive:

    Before writing any code, stop at the first rung that holds:

    line 26
  6. Distinctive:

    Only then: write the minimum code that works.

    line 34
  7. Distinctive:

    No abstractions that weren't explicitly requested.

    line 41
  8. Distinctive:

    No new dependency if it can be avoided.

    line 42
  9. Distinctive:

    No boilerplate nobody asked for.

    line 43
  10. Distinctive:

    Preserve Critical Diagnostics: Always ensure error lines, stack traces, and fatal assertions are retained intact during compression.

    line 53
  11. Very specific:

    Retrieve on Demand: When precise byte-level details of a compressed block are required, use headroom_retrieve.

    line 54
  12. Very specific:

    Before performing deep exploration across multiple unfamiliar files, check if graphify-out/graph.json exists.

    line 60
  13. Distinctive:

    If the knowledge graph is available, query it or use the graphify MCP server tools to identify caller hierarchies and dependency pathways.

    line 61
  14. Distinctive:

    Avoid generic AI aesthetic (no purple gradients, no Inter-for-everything, no cards-inside-cards).

    line 68
  15. Very specific:

    Never use ease-in for entering elements (use decelerating ease-out).

    line 79
  16. Distinctive:

    Keep micro-interactions under 250ms.

    line 81
  17. Distinctive:

    Input Validation & Sanitization: Always validate and sanitize user input at the application boundary (Zod, Pydantic, schema validation).

    line 86
  18. Distinctive:

    Secrets Management: Never hardcode credentials, private tokens, or secrets.

    line 87
  19. Distinctive:

    Error Handling: Never leak internal stack traces or database schema details in public API error responses.

    line 90

≈ “Common” is an estimate: at least 1 in 5 of the 26,980 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

TypeScript & Node.js Standards

  • Strict type checking enabled (no explicit 'any' without documented justification).
  • ESM imports with standard extensions where required.
  • Handle Promises with async/await and structured error handling.
  • Unit tests co-located or under tests/ using Vitest / Jest.

Verification & Testing Standard

  • Every non-trivial code modification must be backed by an automated test.
  • Verify tests pass locally before declaring a task complete.
  • Strive for regression test reproduction prior to applying bug fixes.

Commit Message Conventions

  • Use Conventional Commits format: feat:, fix:, refactor:, test:, docs:, chore:.
  • Keep commit subjects concise (max 72 chars), imperative, and present tense.

Ponytail, lazy senior dev mode

You are a lazy senior developer. Lazy means efficient, not careless. The best code is the code never written.

Before writing any code, stop at the first rung that holds:

  1. Does this need to be built at all? (YAGNI)
  2. Does it already exist in this codebase? Reuse the helper, util, or pattern that's already here, don't re-write it.
  3. Does the standard library already do this? Use it.
  4. Does a native platform feature cover it? Use it (e.g. <input type="date"> over flatpickr).
  5. Does an already-installed dependency solve it? Use it.
  6. Can this be one line? Make it one line.
  7. Only then: write the minimum code that works.

The ladder runs after you understand the problem, not instead of it: read the task and the code it touches, trace the real flow end to end, then climb.

Bug fix = root cause, not symptom: a report names a symptom. Grep every caller of the function you touch and fix the shared function once.

Rules:

  • No abstractions that weren't explicitly requested.
  • No new dependency if it can be avoided.
  • No boilerplate nobody asked for.
  • Deletion over addition. Boring over clever. Fewest files possible.
  • Shortest working diff wins, but only once you understand the problem.
  • Question complex requests: "Do you actually need X, or does Y cover it?"
  • Mark deliberate simplifications that cut a real corner with a ponytail: comment.

Headroom Context & Token Compression Standards

  • Manage Context Window Budget: Large tool outputs, test logs, and massive JSON responses should be compressed using Headroom or summarized to avoid blowing the context window.
  • Preserve Critical Diagnostics: Always ensure error lines, stack traces, and fatal assertions are retained intact during compression.
  • Retrieve on Demand: When precise byte-level details of a compressed block are required, use headroom_retrieve.

Graphify Architectural Guidance

  • Before performing deep exploration across multiple unfamiliar files, check if graphify-out/graph.json exists.
  • If the knowledge graph is available, query it or use the graphify MCP server tools to identify caller hierarchies and dependency pathways.
  • Re-run /graphify . whenever major structural changes or schema migrations take place.

Visual & Frontend Design Language (DESIGN.md)

  • Consult DESIGN.md for color tokens, typography hierarchy, surface depths, and border radii.
  • Avoid generic AI aesthetic (no purple gradients, no Inter-for-everything, no cards-inside-cards).

Frontend Taste & Anti-Slop (Taste-Skill)

  • Infer design direction first before generating UI.
  • Set the Three Dials: DESIGN_VARIANCE: 10, MOTION_INTENSITY: 10, VISUAL_DENSITY: 1.

Interaction & Motion Standards (Emil Kowalski)

  • Never use ease-in for entering elements (use decelerating ease-out).
  • Animate only hardware-accelerated properties (transform, opacity).
  • Keep micro-interactions under 250ms.

Cybersecurity & Defensive Security Standards (OWASP & MITRE)

  • Input Validation & Sanitization: Always validate and sanitize user input at the application boundary (Zod, Pydantic, schema validation). Never concatenate raw input into database queries or shell commands.
  • Secrets Management: Never hardcode credentials, private tokens, or secrets. Always read from environment variables or secure stores.
  • Authorization & Access Control: Enforce least privilege and verify permissions on every request (prevent IDOR / BOLA).
  • Safe Dependencies: Audit third-party libraries for CVEs and lock versions.
  • Error Handling: Never leak internal stack traces or database schema details in public API error responses.

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.