CLAUDE.md examples from real projects

FluidInferenceCLAUDE.md

FluidAudio

FluidInference/FluidAudio/CLAUDE.md

Instructions**: Implementation first, explanation second 2. **Testing Policy**: Add unit tests when writing new code. 3. **Git Operations**: Never run `git push` unless explicitly requested. - **No Co-Author Tags

No other project uses it yet

vitodeployCLAUDE.md

vito

vitodeploy/vito/CLAUDE.md

different user, `cd`/`sudo` must run as the target user, not the login user. ## Code Review - `/review-pr [target-branch]` runs the three project reviewer agents (`php-laravel-reviewer`, `frontend-reviewer

CosmosOSReads credentialsCLAUDE.md

Cosmos

CosmosOS/Cosmos/.taskmaster/CLAUDE.md

session task-master next # Find next available task task-master show # Review task details # During implementation, check in code context into the tasks and subtasks task-master update-subtask

subsurfaceCLAUDE.md

subsurface

subsurface/subsurface/CLAUDE.md

commit messages and PR descriptions. Prefer small, focused commits over large sweeping changes. ## Code Review Rules When reviewing pull requests or commits: - ALWAYS use `git show -- ` to read the actual

SuperCmdLabsCLAUDE.md

SuperCmd

SuperCmdLabs/SuperCmd/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

Chachamaru127Deletes or force-pushesCLAUDE.md

claude-code-harness

Chachamaru127/claude-code-harness/CLAUDE.md

/work" | | breezing | Full parallel run with Agent Teams | "run with team", "breezing" | | harness-review | Code review, quality checks | "review", "security", "performance" | | harness-plan | Planning and task shaping into Plans.md | "plan

plugin87CLAUDE.md

ux-ui-agent-skills

plugin87/ux-ui-agent-skills/CLAUDE.md

behind /critique .claude/skills/ ← Runnable skills (invoke via /name): design-tokens, design-component, design-code, design-review, a11y-audit, apply-aesthetic, redesign, data-dashboard, migrate-design-system, prototype, ux-writing, governance

peteromalletCLAUDE.md

desloppify

peteromallet/desloppify/docs/CLAUDE.md

Claude Code Overlay Use Claude subagents for subjective scoring work. **Do not use `--runner codex`** — use Claude subagents exclusively. ### Review workflow Run `desloppify review --prepare` first to generate review data

coderCLAUDE.md

claudecode.nvim

coder/claudecode.nvim/CLAUDE.md

split. - `keep_terminal_focus` (boolean, default: `false`) - When enabled, keeps focus in the Claude Code terminal when a diff opens instead of moving focus to the diff buffer. This allows

MaxMiksaReads credentialsCLAUDE.md

Auto-Company

MaxMiksa/Auto-Company/CLAUDE.md

Layer | Agent | Persona | When to Use | |-------|------|----------| | `fullstack-dhh` | DHH | Code implementation, technical implementation choices, code review and refactor, dev workflow optimization | | `qa-bach` | James Bach | Test strategy, release quality checks

wordpress-mobileCLAUDE.md

WordPress-Android

wordpress-mobile/WordPress-Android/CLAUDE.md

Kotlin First**: All new classes must be written in Kotlin; avoid Java for new code - **No Empty Lines After Braces**: Do not add empty lines immediately after opening braces (`{`) - **Companion

GitToolsCLAUDE.md

GitVersion

GitTools/GitVersion/CLAUDE.md

completing changes, including scope selection, validation of findings, reviewing fixes, and reporting authentication or service-limit blockers. Install the Claude Code plugin from the repository root with `claude plugin install

KingKongRoboticsCLAUDE.md

jumper

KingKongRobotics/jumper/CLAUDE.md

once, and three tasks were running three different curricula under one name. **Vendored code.** `rl/mjlab/` and `rl/rsl_rl/` are modifiable copies kept word-for-word with upstream so they

MicrosoftCLAUDE.md

mxc

microsoft/mxc/CLAUDE.md

Claude instructions @AGENTS.md For code review, follow [.github/skills/mxc-code-review/SKILL.md](.github/skills/mxc-code-review/SKILL.md

No other project uses it yet

vvaltchevReads credentialsCLAUDE.md

tilck

vvaltchev/tilck/CLAUDE.md

cross). - **Bash shebang**: `#!/usr/bin/env bash`, not `#!/bin/bash` (FreeBSD bash is at `/usr/local/bin/bash`). ## Coding Style **Style matters as much as functionality.** The user rejects a diff that solves the problem

stravuCLAUDE.md

crystal

stravu/crystal/CLAUDE.md

tests with Playwright UI pnpm test:headed # Run tests in headed browser mode ``` ### Code Quality ```bash pnpm typecheck # Type checking across all workspaces pnpm lint # Linting across all workspaces ``` ## Project

chatermCLAUDE.md

Chaterm

chaterm/Chaterm/CLAUDE.md

cover 5. **No Emojis:** Prohibited in code, comments, logs, strings 6. **English Comments:** All code comments must be in English 7. **No Console Logging:** Use `createLogger(module)` from the project

guillermolg00CLAUDE.md

morphicons

guillermolg00/morphicons/CLAUDE.md

export size gates** (size-limit `"import": "{ svgToIcon }"`) — coupling two adapters trips CI, not a reviewer's eye. The aggregate adapters gate may grow as adapters land; core/dom/binding gates never move

No other project uses it yet

Google CloudReads credentialsCLAUDE.md

professional-services / agent-eval

GoogleCloudPlatform/professional-services/tools/agent-eval/CLAUDE.md

them 1. **AI generation** (recommended): `agent-eval init` → "Generate with AI". Gemini analyzes agent code, picks managed metrics, generates 2-4 binary `custom_llm_judge` metrics tailored to the agent

odygrdCLAUDE.md

quill

odygrd/quill/CLAUDE.md

queues. - It orders events by timestamp. - It performs formatting and sink I/O. When reviewing or changing code, always ask: - Is this on the frontend hot path or backend path? - Does