How real projects brief Claude Code: the commands, conventions and warnings they put in CLAUDE.md.
Orchestra-ResearchCLAUDE.mdOrchestra-Research/AI-Research-SKILLs/CLAUDE.md
Trusted Publishers settings. - **Provenance**: `--provenance` flag signs packages with Sigstore for supply chain security. - **Workflow**: `.github/workflows/publish-npm.yml` ### Bumping Versions **Always use `npm version`** (not manual edits) to keep `package-lock.json` in sync
MemTensorCLAUDE.mdMemTensor/MemOS/CLAUDE.md
call chains, gather evidence | | `design-reviewer` | Read-only | Review design docs (architecture / interface / performance / security / requirement coverage) | | `code-reviewer` | Read-only | Review diffs and return APPROVE or CHANGES_REQUESTED | | `backend
JeffallanCLAUDE.mdJeffallan/claude-skills/CLAUDE.md
existing skill directories. **Domain values:** `language` · `backend` · `frontend` · `infrastructure` · `api-architecture` · `quality` · `devops` · `security` · `data-ml` · `platform` · `specialized` · `workflow` --- ### Reference File Standards Reference files follow the [Agent Skills specification
AnthropicCLAUDE.mdanthropics/claude-for-legal/CLAUDE.md
toolset`); MCP and write tools belong to specific subagent leaves. 2. The README's security table and the `agent.yaml` comments must match what the YAML actually grants. Don't claim
AnthropicCLAUDE.mdanthropics/claude-for-legal/litigation-legal/CLAUDE.md
Matters with media / reputational risk | | CISO | [PLACEHOLDER] | | Data incidents, cyber litigation, regulator inquiries on security | | Board litigation / audit committee chair | [PLACEHOLDER] | | Critical matters, disclosure items | ### This counsel **Counsel:** [PLACEHOLDER] **Reports
AnthropicCLAUDE.mdanthropics/claude-for-legal/privacy-legal/CLAUDE.md
DSAR | [PLACEHOLDER] | | | | DPA negotiation | | | | | High-risk PIA | | | | | Regulator contact | — | [GC + you] | Always | | Suspected breach | — | [Security + GC] | Always | --- ## Seed documents | Doc | Location | Reviewed | Notes | |---|---|---|---| | Privacy policy | [PLACEHOLDER] | | | | DPA template | [PLACEHOLDER] | | | | Reference
AnthropicCLAUDE.mdanthropics/claude-for-legal/product-legal/CLAUDE.md
# Product Legal Practice Profile *Written by cold-start on [DATE]. If you see `[PLACEHOLDER
AppleCLAUDE.mdapple/containerization/CLAUDE.md
Linux backend at runtime. **Sandbox env vars.** `CHProcess` and `VirtiofsdProcess` default to the upstream-secure spawn flags. Per-component opt-outs: - `CONTAINERIZATION_NO_CH_SECCOMP=1` — launch cloud-hypervisor with
PydanticCLAUDE.mdpydantic/monty/CLAUDE.md
model-checks` and `cargo test -p monty-fs` on Linux, macOS, and Windows ## Important Security Notice It's ABSOLUTELY CRITICAL that there's no way for code
datahaven-xyzCLAUDE.mddatahaven-xyz/datahaven/CLAUDE.md
with code in this repository. ## Project Overview DataHaven is an EVM-compatible Substrate blockchain secured by EigenLayer. It bridges Ethereum and Substrate ecosystems through: - EigenLayer AVS integration for security - Snowbridge
AnthropicCLAUDE.mdanthropics/defending-code-reference-harness/CLAUDE.md
Claude for Securing Source Code This repo has two halves: - **Interactive skills** (`.claude/skills/`) — read and write files in the repo (no target-code execution, except `/dnr-hunt` / `/dnr-respond`, which
kyegomezCLAUDE.mdkyegomez/swarms/CLAUDE.md
this many times max_loops=1, ) result = moa.run("What are the best practices for securing a Kubernetes cluster?") ``` **When to use:** High-stakes tasks where you want multiple independent perspectives
enricorosCLAUDE.mdenricoros/big-AGI/CLAUDE.md
procedure alias in it resolves to bare `t.procedure` (the cloud layer lives on `dev`) #### Security Considerations - Server-side keys come from env vars only; client-held keys are user-supplied
OsmanticCLAUDE.mdOsmantic/ODS/CLAUDE.md
merges them with base + GPU overlay. Services bind to `127.0.0.1` by default for security. ### Dashboard API FastAPI app in `extensions/services/dashboard-api/` with modular routers (`routers/agents.py`, `features.py`, `privacy.py`, `setup.py`, `updates.py`, `workflows.py`). Uses
deanpetersCLAUDE.mddeanpeters/Product-Manager-Skills/CLAUDE.md
Growth Pyramid, applied through practitioner field material - ✅ **v0.79 Released (May 15, 2026):** Community contributions — security hardening, new meta-skill, bug fix - Added `skills/pm-skill-creator/` — interactive skill for guided skill design
airweave-aiCLAUDE.mdairweave-ai/airweave/CLAUDE.md
path (just `host.com/{endpoint}`) - Use logger from `ctx` (API) or `sync_context` (during sync) - Security: never use `random.*` for security values (ruff S311); use `secrets` module ### Frontend (TypeScript) - TailwindCSS with
op7418CLAUDE.mdop7418/CodePilot/CLAUDE.md
test.ts` - fail-closed：分类器出错 / 空暂存集 / 非 `docs` 判定一律走全量门禁（防「前置失败被静默放行」回归） 高风险 docs（release / security / runtime / provider / permission / DB schema）即便走 docs 快路，也应手动跑 `npm run test`（必要时 build / smoke
czlonkowskiCLAUDE.mdczlonkowski/n8n-skills/CLAUDE.md
user first) - Secret-free/domain-free templates in `assets/`; fresh secrets generated on the box; secure defaults; DNS/ports preflight; Day-2 update/backup/restore - Python Code nodes need task runners in external mode
Q00CLAUDE.mdQ00/ouroboros/CLAUDE.md
introduced by this PR? 5. If scope is split, does immediate user-data or security risk remain? Use these decisions: - Q1 + Q2: **Changes Requested**. - Q3 + Q5: stop and revisit
FlorianBruniauxCLAUDE.mdFlorianBruniaux/claude-code-ultimate-guide/CLAUDE.md
printable summary ├── cowork.md # Cowork redirect page ├── core/ # Architecture, methodologies, releases, known-issues, visual-reference ├── security/ # security-hardening, sandbox-isolation, sandbox-native, production-safety, data-privacy ├── ecosystem/ # ai-ecosystem, mcp-servers
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.
