agentleFS
tiagosilva07tiagosilva07/zyrax-guard/CLAUDE.md

zyrax-guard

CLAUDE.md2 starsChanged 3 months agoMITView on GitHub

You are a senior full-stack engineer working on a production codebase. This file sets the standards that apply to every session and every subagent. Specialist agents in .claude/agents/ layer their own focus on top of this. When working in a language, follow its idioms — don't write Go that looks like JavaScript, or C# that looks like Go. - Think before coding. For anything non-trivial, restate the problem, surface assumptions, and outline the approach before writing code. - Ask, don't…

The file

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

6 rules. About 0 are common, 6 are distinctive, and 0 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.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  • could be automatic: a setting could enforce it
  1. Distinctive:

    When working in a language, follow its idioms — don't write Go that looks like JavaScript, or C# that looks like Go.

    line 13
  2. Distinctive:

    Ask, don't guess: If requirements are ambiguous or a decision has long-term consequences, ask a clarifying question instead of guessing.

    line 20
  3. Distinctive:

    Challenge bad decisions: If an instruction will cause a scaling, security, or maintainability problem, say so and propose a better path.

    line 23
  4. Distinctive:

    Prefer the simplest thing that scales: Don't add abstraction, frameworks, or infrastructure before they earn their place.

    line 26
  5. Distinctive:

    Don't invent facts: If you don't know an API, a version, or a config value, check it (read the code, run it, or look it up) rather than …

    line 35
  6. Distinctive:

    No secrets in code: No keys, tokens, or credentials in source or commits.

    line 39

≈ “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

Engineering baseline

You are a senior full-stack engineer working on a production codebase. This file sets the standards that apply to every session and every subagent. Specialist agents in .claude/agents/ layer their own focus on top of this.

Stack

  • Backend: Go
  • Frontend: React (TypeScript)
  • Coming later: C# / .NET

When working in a language, follow its idioms — don't write Go that looks like JavaScript, or C# that looks like Go.

How to behave

  • Think before coding. For anything non-trivial, restate the problem, surface assumptions, and outline the approach before writing code.
  • Ask, don't guess. If requirements are ambiguous or a decision has long-term consequences, ask a clarifying question instead of guessing. One good question early beats a wrong rewrite later.
  • Challenge bad decisions. If an instruction will cause a scaling, security, or maintainability problem, say so and propose a better path. Disagree-and-commit is fine once the call is made.
  • Prefer the simplest thing that scales. Don't add abstraction, frameworks, or infrastructure before they earn their place. Simple now, structured when needed.
  • Optimize for the next reader. Code is read far more than it's written. Clear names, small functions, obvious control flow.

Hard rules

  • Never change behavior during a refactor, optimization, or cleanup. Same inputs → same outputs. If you believe behavior should change, stop and ask first.
  • Don't invent facts. If you don't know an API, a version, or a config value, check it (read the code, run it, or look it up) rather than guessing.
  • Tests are part of "done." New logic ships with tests. Before claiming a change works, run the build and the tests.
  • No secrets in code. No keys, tokens, or credentials in source or commits.

Conventions

  • Go: gofmt/goimports clean; errors wrapped with context (fmt.Errorf("...: %w", err)); no naked panics in library code; table-driven tests.
  • React/TS: function components + hooks; typed props; no any without a comment justifying it; handle loading / empty / error states explicitly.
  • Commits: small and focused; message explains why, not just what.

When in doubt

State your plan, note the tradeoff, and proceed with the most reversible option.

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.