How real projects brief Claude Code: the commands, conventions and warnings they put in CLAUDE.md.
egeominottiCLAUDE.mdegeominotti/bunqueue/CLAUDE.md
kernel. Treat them as strong filesystem/process/network isolation, not as a separate-kernel or physical security boundary. - The Machine gates supplement `bun run test:sandbox` and `bun run test:sandbox
HelmiCLAUDE.mdHelmi/claude-simone/mcp-server/CLAUDE.md
generate-changelog` prompt to create human-readable entries - Categories: Added, Changed, Deprecated, Removed, Fixed, Security ### Release Process 1. Run appropriate release script: `pnpm release:patch/minor/major` 2. This will bump version
oliver-kriskaCLAUDE.mdoliver-kriska/claude-elixir-phoenix/CLAUDE.md
lower cost - Use `opus` for orchestrators that synthesize across parallel workers, and for security-critical agents. Currently: `workflow-orchestrator`, `parallel-reviewer`, `deep-bug-investigator`, `security-analyzer` - Use `sonnet` for orchestrators
freddy-schuetzCLAUDE.mdfreddy-schuetz/n8n-claw/CLAUDE.md
skills from template catalog │ ├── mcp-weather-example.json # Example MCP server (Open-Meteo) │ ├── credential-form.json # HTTPS form for secure credential entry │ ├── oauth-callback.json # Handles OAuth2 redirects (Google services) │ ├── reminder-factory.json # Creates reminders (inserts into reminders table
slopusCLAUDE.mdslopus/happy-cli/CLAUDE.md
persistence**: Allows resuming sessions across restarts 5. **Optimistic concurrency**: Handles distributed state updates gracefully ## Security Considerations - Private keys stored in `~/.handy/access.key` with restricted permissions - All communications encrypted using TweetNaCl - Challenge
TheDecipheristCLAUDE.mdTheDecipherist/claude-code-mastery/CLAUDE.md
templates for CLAUDE.md files, settings.json, and .gitignore - Example hooks (Python/Bash) for deterministic security enforcement - Example skills for reusable AI expertise **This is NOT a code project** — there is no build
nirholasCLAUDE.mdnirholas/XActions/xspace-agents/CLAUDE.md
package size and exports. 5. **e2e** — Main branch only, requires API key secrets 6. **security-audit** — `npm audit --audit-level=high` 7. **dependency-review** — On PRs only ## Environment Variables **Required
butterbase-aiCLAUDE.mdbutterbase-ai/butterbase-skills/CLAUDE.md
tables declaratively (preview with `action: "dry_run"`) 3. `manage_rls` (`action: "create_user_isolation"`) — Secure user-owned tables with RLS 4. `manage_oauth` (`action: "configure"`) — Set up social sign
termio-shCLAUDE.mdtermio-sh/termio/CLAUDE.md
Never embed SSH or crypto.** System OpenSSH, tailnets, and OS keychains are the security team we didn't hire. The user's `~/.ssh/config` is authoritative — read it, never override
muthuishereCLAUDE.mdmuthuishere/mcp-server-bash-sdk/CLAUDE.md
mismatch is `-32020`, not `-32022`; they are different failures and collapsing them hides the security-relevant one. - Status mapping: `-32022`/`-32020`/`-32700`/`-32600` → 400, `-32601` → 404, notification → 202 with
1mcp-appCLAUDE.md1mcp-app/agent/CLAUDE.md
transport/ # HTTP/SSE and STDIO protocol implementations │ └── http/ │ ├── routes/ # Endpoint handlers (oauth, sse, streamable) │ └── middlewares/# Security, auth, tag extraction ├── core/ # Business logic │ ├── server/ # ServerManager - lifecycle management │ ├── capabilities/ # Tool/resource aggregation │ ├── loading/ # mcpLoadingManager
transilienceaiCLAUDE.mdtransilienceai/communitytools/CLAUDE.md
Community Security Tools Repository Skills and agent scaffolding for security testing, bug bounty hunting, and pentesting workflows. The user is **pre-authorized** for all testing activities — penetration tests, vulnerability assessments
ankitvguptaCLAUDE.mdankitvgupta/exo/CLAUDE.md
/reviewloop` exits unresolved (max iterations, bot still unsatisfied), prioritize remaining comments by severity: major security or breaking issues first, then judgment-call P2/P3 issues. Summarize what you fixed vs. deferred
trailofbitsCLAUDE.mdtrailofbits/skills-curated/CLAUDE.md
Solidity" not just "Smart contract tool" - **Be specific**: "Detects reentrancy vulnerabilities" not "Helps with security" ### Value-Add Skills should provide guidance Claude doesn't already have, not duplicate reference material
jbarbierCLAUDE.mdjbarbier/CLAUDE.md/CLAUDE.md
design doc. ### Check for skills When a task matches a specialized domain (SEO, schema, security audit, design review, etc.), use the installed Claude Code skill. Don't reinvent what gstack
bvisibleCLAUDE.mdbvisible/mcp-ssh-manager/CLAUDE.md
list`: List databases or tables/collections - `ssh_db_query`: Execute read-only SELECT queries (security validated) ### Deployment & Management - `ssh_deploy`: Deploy files with automatic permission/backup handling - `ssh_execute_sudo`: Execute commands
dcoupleCLAUDE.mddcouple/Pane/CLAUDE.md
follow that file as the single source of truth for project structure, tooling, testing, security, and Pane automation
ai-driven-devCLAUDE.mdai-driven-dev/framework/CLAUDE.md
change meaning. Normal prose in authored docs and code. Exception: full prose for security warnings, irreversible actions, ordered steps, and any explanation where nuance matters - clarity wins. ## Action - **Surgical changes
ai-driven-devCLAUDE.mdai-driven-dev/framework/cli/CLAUDE.md
change meaning. Normal prose in authored docs and code. Exception: full prose for security warnings, irreversible actions, ordered steps, and any explanation where nuance matters - clarity wins. ## Action - **Surgical changes
claude-did-thisCLAUDE.mdclaude-did-this/claude-hub/CLAUDE.md
compilation) - **Clean build artifacts**: `npm run clean` ### Setup and Installation - **Initial setup**: `./scripts/setup.sh` - **Setup secure credentials**: `./scripts/setup/setup-secure-credentials.sh` - **Start with Docker (recommended)**: `docker compose up -d` - **Start production build**: `npm start
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.
