How real projects brief Claude Code: the commands, conventions and warnings they put in CLAUDE.md.
NVIDIACLAUDE.mdNVIDIA/aicr/.claude/CLAUDE.md
with error codes (never `fmt.Errorf`) 7. **Context timeouts** — All I/O operations need context with timeout 8. **Check context in loops** — Always check `ctx.Done()` in long-running operations ## Review Output Links
ZingerLittleBeeCLAUDE.mdZingerLittleBee/Heeler/CLAUDE.md
Sibling deliverables live in this repo: `plugin/` is the herdr plugin that renders Pairing Codes and posts Agent Notifications (Node, zero framework, `npm test`); `relay/` is the stateless Push Relay
ChenChenyaqiCLAUDE.mdChenChenyaqi/learn-anything/CLAUDE.md
coding assistants, turning them into interactive learning tutors. It supports 30+ AI tools (Claude Code, Cursor, Gemini CLI, Codex, Copilot, Windsurf, etc.) and outputs localized files
notqueCLAUDE.mdnotque/vexjoy-agent/CLAUDE.md
frontmatter checking), use a script. Reserve LLM judgment for contextual diagnosis, design decisions, and code review. **Write dense.** High fidelity, minimum words. Prefer tables and lists over paragraphs
cwinvestmentsCLAUDE.mdcwinvestments/memstack/CLAUDE.md
posts, emails, ads, landing pages) - Any project planning or architecture decisions - Any refactoring or code review - Any task where the user asks "how should I..." or "what's the best
vivekchandCLAUDE.mdvivekchand/clawmetry/CLAUDE.md
goals, alternatives rejected, risk accepted -- then the blueprint, then the code. A requirement written after the fact reviews nothing, and `scripts/check_product_record.py` gates PRs on citing one (or an explicit
alirezarezvaniCLAUDE.mdalirezarezvani/ClaudeForge/agent/CLAUDE.md
model: haiku` — token efficiency for routine sync runs. - `color: purple` — visual indicator in Claude Code UI. - `fork_safe: true` — agent runs independently. - `hooks: [SessionStart, PreToolUse, PostToolUse]`. 3. Workflow phases inside
alirezarezvaniCLAUDE.mdalirezarezvani/ClaudeForge/CLAUDE.md
auto-imports these): @skill/CLAUDE.md @command/CLAUDE.md @agent/CLAUDE.md @docs/CLAUDE.md ## Behavioral Guidelines Behavioural guardrails applied to every coding, review, and refactoring task in this repo. Full skill: `~/.claude/skills/karpathy-guidelines/SKILL.md`. 1. **Think before coding.** State
1mancompanyCLAUDE.md1mancompany/OneManCompany/CLAUDE.md
deploy, push, create PR → invoke ship - QA, test the site, find bugs → invoke qa - Code review, check my diff → invoke review - Update docs after shipping → invoke document-release - Weekly retro
AspegioCLAUDE.mdAspegio/nelson/CLAUDE.md
next regression. If you find existing naked suppressions, treat them as a mini code-review task: add a reason or remove the suppression. If a threshold (`max-complexity`, `max-args
bassimeledathCLAUDE.mdbassimeledath/dispatch/CLAUDE.md
Dispatch - Development Guide Dispatch is a skill (`/dispatch`) for Claude Code that decomposes large coding tasks into subtasks, dispatches them to background AI worker agents, and tracks progress via checklist
Community-AccessCLAUDE.mdCommunity-Access/accessibility-agents/claude-code-plugin/CLAUDE.md
Development This project enforces WCAG AA accessibility standards for all web UI code. ## Hook-Based Enforcement Accessibility review is enforced by three global hooks: 1. **Proactive detection** (`UserPromptSubmit`) — Detects
Community-AccessCLAUDE.mdCommunity-Access/accessibility-agents/CLAUDE.md
Development This workspace enforces WCAG AA accessibility standards for all web UI code. ## Hook-Based Enforcement Accessibility review is enforced by three global hooks, not just instructions: 1. **Proactive detection
dsifryCLAUDE.mddsifry/metaswarm/CLAUDE.md
driven development** (`superpowers:subagent-driven-development`) — Dispatch subagents per task in this session with code review between tasks. Faster, lighter-weight, lower token cost. > 3. **Parallel session** (`superpowers:executing-plans
gtmagentsCLAUDE.mdgtmagents/gtm-agents/CLAUDE.md
these automatically. --- ## Architecture Principles ### Business-First Design - Designed for non-technical GTM users - No coding required - Natural language interaction supported ### Token Optimization - Average 3.2 components per plugin (follows Anthropic
MicrosoftCLAUDE.mdmicrosoft/semanticworkbench/assistants/document-assistant/CLAUDE.md
Optional` from typing, using the `| `syntax. - Always add appropriate type hints such that the code would pass a pywright type check. - Do not add extra newlines before loops. - For type
MicrosoftCLAUDE.mdmicrosoft/semanticworkbench/assistants/knowledge-transfer-assistant/CLAUDE.md
Trust in emergence from simple, well-defined components - Direct library integration with minimal wrappers ### Code Style - Use 4 spaces for indentation - Maximum line length: 120 characters - Follow PEP 8 naming
MicrosoftCLAUDE.mdmicrosoft/semanticworkbench/CLAUDE.md
templates * **AI tools**: Provide relevant context files for better code generation and debugging * **Code reviews**: Reference context files to understand cross-component impacts ## Common Commands * Build/Install: `make install` (recursive
ruilisiCLAUDE.mdruilisi/lsbot/CLAUDE.md
refs to click/type/drag/select 3. Re-snapshot if the page changed ## Skills ### /review-code - Code Review After Changes Review code changes in the current branch compared to main. Provide structured feedback
LarsCoweCLAUDE.mdLarsCowe/bmalph/CLAUDE.md
drivers/ # Platform driver scripts (claude-code, codex, opencode, copilot, cursor) │ ├── lib/ # Shell libraries (circuit breaker, response analysis, etc.) │ └── templates/ # Prompt, agent, fix plan, review, and ralphrc templates ├── bmalph/ # bmalph state
A file Claude Code reads at the start of every session. It holds the commands, conventions and warnings the agent needs for this project.
At the repository root. Claude Code also reads CLAUDE.md files in subdirectories when it works there.
Build and test commands, the project's layout, conventions that aren't obvious from the code, and mistakes to avoid. Short files tend to work better than long ones.
Claude Code reads CLAUDE.md; most other agents read AGENTS.md. Many projects keep one and point the other at it.
