CLAUDE.md examples from real projects

silversteinCLAUDE.md

minutes

silverstein/minutes/CLAUDE.md

capture + streaming) │ │ ├── transcribe.rs # Batch transcription: whisper.cpp. Retained Parakeet config resolves to Whisper until secure transport exists │ │ ├── diarize.rs # Speaker diarization + attribution types (pyannote-rs, or energy-based from per-source stems

No other project uses it yet

tom-james-watsonCLAUDE.md

breaktimer-app

tom-james-watson/breaktimer-app/CLAUDE.md

Settings, etc.) - **Styling**: CSS with Tailwind CSS - **Sounds**: Audio files for notifications - **Preload**: `preload.js` - Secure IPC bridge - **Fonts**: Inter font bundled in `public/fonts/` ### Types (`app/types/`) - Shared TypeScript definitions

nats-ioCLAUDE.md

nats.rs

nats-io/nats.rs/CLAUDE.md

fully async, Tokio-based client. > **The `nats` crate (synchronous client) is deprecated.** It receives security fixes only. All new > work happens in `async-nats`. Do not modify the `nats/` directory

No other project uses it yet

six2dezCLAUDE.md

burp-ai-agent

six2dez/burp-ai-agent/CLAUDE.md

exposes Burp operations to external AI agents over the Model Context Protocol (MCP). Targets security professionals running Burp Community or Professional on macOS / Linux / Windows. **Core Value:** Bring modern

LUC4N3XCLAUDE.md

Levyra-deepsound / claude

LUC4N3X/Levyra-deepsound/.agents/claude/CLAUDE.md

levyra-compose`, `levyra-design-taste`, `levyra-android-performance`, `levyra-r8-proguard`, `levyra-android-intent-security`, `levyra-ci-workflows`, `levyra-context-efficiency`, `levyra-pr-review`, `levyra-release-check`. The shared router

Open-Source-LegalCLAUDE.md

OpenContracts

Open-Source-Legal/OpenContracts/CLAUDE.md

issue #820): operational exceptions are caught and returned as error strings to the LLM; security exceptions (`PermissionError`, `ToolConfirmationRequired`) propagate. - Pre-execution checks run on every call (not cached): permission validation

kaansenol5CLAUDE.md

VibeOS

kaansenol5/VibeOS/CLAUDE.md

mountable on macOS - [x] Interrupts - GIC-400 working! Keyboard via IRQ, boots at EL3 (Secure) - [x] Timer - 10ms tick (100Hz), used for uptime tracking - [x] System Monitor - GUI app showing

daymadePipes a download into a shellCLAUDE.md

claude-code-skills

daymade/claude-code-skills/CLAUDE.md

commands from another working directory, use the [fixed command entry](daymade-skill/skill-creator/references/fixed-command-entry.md). For security scans, use its [absolute owner-script recipe](daymade-skill/skill-creator/references/publishing-and-packaging.md#step-6-security-review). Before

umputunReads credentialsCLAUDE.md

ralphex

umputun/ralphex/CLAUDE.md

# ralphex Autonomous plan execution with Claude Code - Go rewrite of ralph.py. ## LLM Documentation See

geodroReads credentialsCLAUDE.md

lerd

geodro/lerd/CLAUDE.md

report issues use `Refs #N` and stay open until the stable release ships — **except** security issues, which close once the fix merges to main. - PR/issue comment style: casual plain prose

ishaan1013CLAUDE.md

shadow

ishaan1013/shadow/CLAUDE.md

task management. The platform provides both local and remote execution modes with enterprise-grade security. ### Core Architecture **Monorepo Structure** (Turborepo managed): - `apps/frontend/` - Next.js 15 + React 19 application with real-time

ducktorsCLAUDE.md

turborepo-remote-cache

ducktors/turborepo-remote-cache/CLAUDE.md

symbol — callers, callees, which execution flows it participates in — use `context({name: "symbolName"})`. - For security review, `explain({target: "fileOrSymbol"})` lists taint findings (source→sink flows; needs `analyze --pdg`). ## Never

amir20CLAUDE.md

dtop

amir20/dtop/CLAUDE.md

# CLAUDE.md This file provides guidance to Claude Code (claude.ai/code) when working with code

psd-toolsCLAUDE.md

psd-tools

psd-tools/psd-tools/CLAUDE.md

public surface: a new or renamed symbol, a new parameter, a corrected annotation - `[security]` — security fixes - `[docs]` — documentation only - `[ci]` — what a user installs: the wheels, platforms or sdist contents

No other project uses it yet

akitaonrailsReads credentialsCLAUDE.md

ai-jail

akitaonrails/ai-jail/CLAUDE.md

migration inputs. ai-jail is a developer-friendly accident guard, not a hard-security or malware/invasion boundary. It defends against a trusted-but-fallible agent's mistakes (a mistyped command

kizuna-ai-labCLAUDE.md

sokuji

kizuna-ai-lab/sokuji/CLAUDE.md

# CLAUDE.md This file provides guidance to Claude Code (claude.ai/code) when working with code

win4rCLAUDE.md

ClawTeam-OpenClaw

win4r/ClawTeam-OpenClaw/CLAUDE.md

DATA_DIR` relocates team/task/inbox state but NOT global config. ## Cross-cutting conventions ### Path discipline (security-critical) Every user-supplied name (team, agent, user, plan ID, inbox recipient…) must flow through

rmusser01Reads credentialsCLAUDE.md

tldw_server

rmusser01/tldw_server/CLAUDE.md

appropriate HTTP status codes - Provide meaningful error messages - Log errors with context using loguru ### Security Best Practices - Validate all user input - Use parameterized queries for database operations - Never log sensitive

ghostwrightReads credentialsCLAUDE.md

phantom

ghostwright/phantom/CLAUDE.md

docker-entrypoint.sh # Docker bootstrap (wait for deps, model pull, init) docs/ # Documentation (architecture, channels, mcp, security, etc.) ``` ## Architecture Overview Message flow: Slack message -> SlackChannel adapter -> ChannelRouter -> SessionManager (find/create session) -> PromptAssembler (base

jekalminDeletes or force-pushesCLAUDE.md

extended_openai_conversation

jekalmin/extended_openai_conversation/CLAUDE.md

image` (vision API), `change_config` (runtime config updates), `reload_skills`, `download_skill` (from GitHub). ### Security - Workspace restriction for file operations (default: `extended_openai_conversation/`) - Shell command deny patterns