How real projects brief Claude Code: the commands, conventions and warnings they put in CLAUDE.md.
agentic-communityCLAUDE.mdagentic-community/mcp-gateway-registry/CLAUDE.md
Provide usage examples for complex functions - Keep docstrings up-to-date with code changes ## Security Guidelines ### Security invariants (always apply) These are hard rules distilled from prior security-review findings
kossakovskyCLAUDE.mdkossakovsky/selfhost-ai/CLAUDE.md
scripts via `source "$(dirname "$0")/utils.sh" && init_paths`) - `scripts/01_system_preparation.sh`: System updates, firewall, security hardening - `scripts/02_install_docker.sh`: Docker and Docker Compose installation - `scripts/git.sh`: Git utilities (sync with origin, branch detection, configuration) - `scripts/03_generate_secrets.sh
apifyCLAUDE.mdapify/mcpc/CLAUDE.md
agents. `mcpc` can connect to any MCP server over Streamable HTTP or stdio transports, securely login via OAuth credentials and store credentials, and keep long-term sessions to multiple servers
QuantGeekDevCLAUDE.mdQuantGeekDev/mcp-framework/CLAUDE.md
minutes (configurable) - Allows immediate token revocation - Slower: ~20-50ms per request (cached) **Security Features:** - Tokens must be in Authorization header (Bearer scheme) - Tokens in query strings rejected automatically (security
opentabs-devCLAUDE.mdopentabs-dev/opentabs/CLAUDE.md
adapter IIFE source code and a review token → AI reviews the code with security guidance → user confirms → call `plugin_mark_reviewed` with the token to set the permission and mark
JohnRiceMLCLAUDE.mdJohnRiceML/clawport-ui/CLAUDE.md
parser (`parseSlashCommand`), matcher (`matchCommands`), executor (`executeCommand`) | | `lib/id.ts` | `generateId()` -- UUID generator with fallback for non-secure contexts (HTTP, older browsers) | | `lib/pipeline-utils.ts` | Pipeline layout builder, health check prompt builder, cron context builder
MicrosoftCLAUDE.mdmicrosoft/fabric-toolbox/tools/MicrosoftFabricMgmt/CLAUDE.md
Level Warning } ``` **Example output** for a 400 with `moreDetails[0] = { errorCode: "UniversalSecurityFeatureDisabledForWorkspace", message: "Universal security feature is disabled for the workspace" }`: - **Debug**: `Failed to retrieve... Full error: API request failed
MicrosoftCLAUDE.mdmicrosoft/fabric-toolbox/tools/MicrosoftFabricMgmtMCPServer/CLAUDE.md
catch` block in the sentinel wrapper as `"error_type"` = PS exception class name) --- ## Security Constraints - **Never use `shell=True`** in `subprocess.Popen` — use a list of arguments. - **App secrets** passed
pocketpawCLAUDE.mdpocketpaw/pocketpaw/CLAUDE.md
screenshots). `BrowserDriver` returns `NavigationResult` with a `refmap` mapping ref numbers to CSS selectors - **Security** (`security/`) — Guardian AI (secondary LLM safety check) + append-only audit log (`~/.pocketpaw/audit.jsonl`) - **Tools** (`tools/`) — `ToolProtocol` with
pocketpawCLAUDE.mdpocketpaw/pocketpaw/docs/CLAUDE.md
start, config, project structure ├── concepts/ # 6 pages — architecture, message bus, agent loop, memory, tools, security ├── channels/ # 10 pages — overview + 9 channel guides ├── backends/ # 8 pages — overview + 6 backend guides + Ollama
jdevalkCLAUDE.mdjdevalk/specification.website/CLAUDE.md
citation with the standard. 2. **Stay platform-agnostic.** Describe outcomes, not implementations. "Set `Content-Security-Policy`" is in scope. "Add this to your `next.config.mjs`" is not. Link out to platform
LeslieLeungCLAUDE.mdLeslieLeung/glean/CLAUDE.md
pnpm build:linux # Linux (AppImage + deb) ``` Built applications will be in `frontend/apps/web/release/`. ### Key Features - **Secure token storage**: Uses `electron-store` for encrypted credential storage - **Auto-updates**: Built-in update checker
alirezarezvaniCLAUDE.mdalirezarezvani/claude-code-skill-factory/documentation/CLAUDE.md
agent types **Example Agents**: - Code reviewer - Frontend developer - Backend developer - Test automation specialist - Security engineer --- ### PROMPTS_FACTORY_PROMPT.md **Purpose**: Generate domain-specific prompt builders (complete prompt generation systems) **Use for**: Creating specialized
alirezarezvaniCLAUDE.mdalirezarezvani/claude-code-skill-factory/generated-skills/CLAUDE.md
Cases**: - Automate M365 tenant setup - Manage user lifecycle operations - Generate PowerShell automation scripts - Configure security and compliance settings - Bulk user provisioning and management --- ### 4. Psychology Advisor (31 KB) **Location**: `psychology
borgheiCLAUDE.mdborghei/Claude-Skills/CLAUDE.md
markdown→HTML publishing (stdlib only, zero network) | | **Standards Library** | [standards/CLAUDE.md](standards/CLAUDE.md) | Communication, quality, git, security standards | | **Templates** | [templates/CLAUDE.md](templates/CLAUDE.md) | Template system usage | ## Architecture Overview ### Repository Structure ``` claude-code-skills/ ├── .claude
borgheiCLAUDE.mdborghei/Claude-Skills/c-level-advisor/CLAUDE.md
sales, GTM - **coo-advisor/** — operations, scaling, process - **chro-advisor/** — talent, culture, comp - **ciso-advisor/** — security strategy, risk, compliance - Plus: board-deck-builder, board-meeting, change-management, chief-of-staff, company
borgheiCLAUDE.mdborghei/Claude-Skills/documents/CLAUDE.md
document audit (comments, tracked changes, headings, styles) | | PDF Toolkit | `pdf-toolkit/` | `pdf_auditor.py` | PDF metadata, security, encryption, embedded JS / files | | Pptx Toolkit | `pptx-toolkit/` | `pptx_auditor.py` | PowerPoint audit (slides, density, notes, animations
borgheiCLAUDE.mdborghei/Claude-Skills/markdown-html/CLAUDE.md
HTML-escaped at the boundary. Treat any change that emits unescaped input as a security regression. ## Common Patterns ``` markdown-html/ / ├── SKILL.md ├── references/ │ └── *.md (rulebooks, severity models, layout and density guidance
borgheiCLAUDE.mdborghei/Claude-Skills/project-management/CLAUDE.md
agent-ready tickets, human accountability, review-capacity gate, DORA + agent metrics by author type, security, rollout (2 tools: `agent_delegation_scorer.py`, `agent_delivery_metrics.py`) ### Role-Based Skills (10) 1. **senior-pm/** - Portfolio management, stakeholder
borgheiCLAUDE.mdborghei/Claude-Skills/ra-qm-team/CLAUDE.md
This guide covers the **27 production-ready compliance skills** spanning medical device regulations, information security, privacy, financial resilience, AI governance, and infrastructure security — plus the new `audit-prep/` subfolder with
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.
