CLAUDE.md examples from real projects

buxukuCLAUDE.md

SmartSub

buxuku/SmartSub/CLAUDE.md

Claude Project Memory Template **Primary Directive**: "Evidence > assumptions | Code > documentation | Efficiency > verbosity" ## Project Context - **Current Branch**: `[BRANCH_NAME]` (if applicable) - **Main Branch**: `[DEFAULT_BRANCH]` (for PRs) - **Active Work**: `[CURRENT_FOCUS

Galaxy-DawnCLAUDE.md

claude-scholar

Galaxy-Dawn/claude-scholar/CLAUDE.md

assumption briefly. --- ## Execution Priorities - Check facts before making claims. - Verify after changing files, code, documentation, or configuration. - Keep changes small, reversible, and easy to review. - Confirm before destructive or high

areal-projectCLAUDE.md

AReaL

areal-project/AReaL/CLAUDE.md

body - **Squash**: Squash WIP commits before opening PR - **PR requirements**: Run pre-commit, document test coverage, note hardware limitations ## Extended Configuration See `.claude/agents/`, `.claude/skills/`, `.claude/commands/`, and `.claude/rules/` for specialized instructions

kurrent-ioCLAUDE.md

KurrentDB

kurrent-io/KurrentDB/CLAUDE.md

Microsoft Docs**: `microsoft_docs_search`, `microsoft_docs_fetch` — query for .NET/Azure docs - **Context7**: Library documentation and code examples - Config: `.claude/settings.json` (committed), `.claude/settings.local.json` (local) ## Configuration Defaults (v25.1) - **SecondaryIndexing:Enabled** — `true` - **ServerGC

vikibossCLAUDE.md

60s

vikiboss/60s/CLAUDE.md

responses - **Validation**: Validate required parameters using `Common.requireArguments()` - **Consistency**: Follow existing code patterns in `src/modules/` - **Documentation**: Add JSDoc comments for complex functions - **Testing**: Test all three encoding formats (json/text/markdown) when adding

pytorchCLAUDE.md

torchtitan

pytorch/torchtitan/.claude/CLAUDE.md

shared by reference. Be explicit when deep copies are needed. - See `.claude/rules/config.md`. ### Comments and Documentation - Add comments only for genuinely non-obvious things: dimension semantics, parallelism gradient placements

coleam00CLAUDE.md

ottomator-agents / agentic-rag-knowledge-graph

coleam00/ottomator-agents/agentic-rag-knowledge-graph/CLAUDE.md

python commands with 'python3' ### 🔌 MCP Server Usage #### Crawl4AI RAG MCP Server - **Use for external documentation**: Get docs for Pydantic AI - **Always check available sources first**: Use `get_available_sources

coleam00CLAUDE.md

ottomator-agents / claude-agent-sdk-demos

coleam00/ottomator-agents/claude-agent-sdk-demos/CLAUDE.md

NEVER code without checking current tasks first.** ## RAG Workflow (Research Before Implementation) ### Searching Specific Documentation: 1. **Get sources** → `rag_get_available_sources()` - Returns list with id, title, url 2. **Find

coleam00CLAUDE.md

ottomator-agents / claude-skill-archon

coleam00/ottomator-agents/claude-skill-archon/CLAUDE.md

available across all projects ## Repository Structure ``` claude_code_skills/ ├── README.md # Complete workshop guide and documentation ├── .claude/ │ ├── skills/ │ │ └── archon/ # Example skill: Archon integration │ │ ├── SKILL.md # Main skill definition with YAML frontmatter │ │ ├── scripts

coleam00Reads credentialsCLAUDE.md

ottomator-agents / livekit-agent

coleam00/ottomator-agents/livekit-agent/CLAUDE.md

mode=auto` - **Mock external services** in unit tests - **Use behavioral testing** for conversation flows ## 📚 Documentation Structure When documenting agents: 1. **Purpose**: What the agent does 2. **Configuration**: Required environment variables

shader-slangCLAUDE.md

slang

shader-slang/slang/CLAUDE.md

access by role/key, not by position. - **Keep a working log/report.** Maintain a scratch markdown document throughout the task that records: the problem and a motivating example, road-blockers encountered

shader-slangCLAUDE.md

slang / extras

shader-slang/slang/extras/claude.md

reviewer: checks correctness, style, edge cases, test coverage, and regressions. It also reviews all documentation changes and reverts additions that are not significant enough to justify a permanent change

steveruizokCLAUDE.md

perfect-freehand

steveruizok/perfect-freehand/CLAUDE.md

with Prettier yarn format # Check formatting without making changes yarn format:check # Generate API documentation cd packages/perfect-freehand && yarn docs ``` ## Architecture This is a **yarn workspaces monorepo** with two packages: - `packages/perfect-freehand

No other project uses it yet

middleapiCLAUDE.md

orpc

middleapi/orpc/CLAUDE.md

type — never `feat`/`fix`/`perf`/`docs`, since these changes do not affect library users. Documentation about a package or protocol uses that package/protocol scope instead (e.g. `docs(openapi): ...`). - Omit

VercelCLAUDE.md

streamdown / website

vercel/streamdown/apps/website/.claude/CLAUDE.md

boundary conditions and error states 5. **User experience** - Accessibility, performance, and usability considerations 6. **Documentation** - Add comments for complex logic, but prefer self-documenting code --- Most formatting and common issues

No other project uses it yet

SenteLabsAIReads credentialsCLAUDE.md

OpenExecutive

SenteLabsAI/OpenExecutive/CLAUDE.md

also gitignored. Structure: - `company/profile.yaml` — structured company profile (populated by onboarding wizard) - `company/docs/` — uploaded documents (indexed into ChromaDB) ## Code Style - Python: `ruff` for linting, `mypy` for type checking, `pytest` for tests

umputunCLAUDE.md

remark42

umputun/remark42/CLAUDE.md

with a type and an area (add priority when relevant): - Type: `bug`, `enhancement`, `question`, `documentation`, `discussion` - Area: `backend`, `frontend`, `site`, `CI`, `design`, `localization` - Priority: `important`, `minor`, `some day` - Contribution: `help

umputunCLAUDE.md

remark42 / frontend

umputun/remark42/frontend/CLAUDE.md

those patch floors into `engines` or the docs, or every lockfile refresh becomes a documentation change. ## One manifest, at `frontend/apps/remark42` `frontend/` holds no `package.json`, no lockfile and no `pnpm-workspace.yaml`. Everything

phanCLAUDE.md

phan

phan/phan/CLAUDE.md

issue emission in appropriate visitor 3. Add test cases in `tests/files/` 4. Update documentation ### Creating a Plugin 1. Create plugin in `.phan/plugins/` 2. Extend `\Phan\PluginV3` 3. Implement required visitor

No other project uses it yet

zenml-ioCLAUDE.md

zenml

zenml-io/zenml/CLAUDE.md

ZenML Claude Code Guidelines This document provides guidance for Claude Code when working with the ZenML codebase. ZenML is an extensible, open-source MLOps framework for creating production-ready