What is crow CLAUDE.md?

Repository instructions for working with Crow, including rules for commits, rebasing before pushes, tests, server startup, and maintaining instruction files. A repository is the project’s tracked source and configuration files.

Use them when changing Crow, committing files, pushing branches, running the test suite, starting servers, or updating CLAUDE.md and crow.md.

They give coding agents the project-specific rules needed to avoid sweeping in unrelated work and to run the expected checks correctly.

crow CLAUDE.md — Instructions file by kh0pper

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods

npx agentmods add instructions/kh0pper/crow/claude-md

Clone the repo

git clone --depth 1 https://github.com/kh0pper/crow

Per session 2,090 This file is loaded in full into every session.

When invoked 2,090 The same file — it is already loaded in full.

Security scan A 0 findings. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5	$0.02090	$0.02090
Opus 5	$0.01045	$0.01045
Sonnet 5	$0.00418	$0.00418
Haiku 4.5	$0.00209	$0.00209

Measured 2d ago against content hash 416ce7fbfa45, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

crow CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

CLAUDE.md · 63 lines

How it starts

The opening of the file, as written. The whole thing — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

Index for working on this repo. Load-bearing rules inline; deep reference is in docs/.

Working with this repo

Always commit with a positional path arg: git commit <path> -m "...", not git add <path> && git commit -m "...". Parallel Claude sessions modify the working tree concurrently; git add then a bare git commit will sweep in unrelated WIP. Verify with git show --stat HEAD after every commit. (See ~/.claude/CLAUDE.md Learnings 2026-04-14.)
Always git pull --rebase before pushing a branch — parallel sessions commonly push to main between your fetch and your push.
Tests: npm test runs the FULL suite in a scratch env (safe on any box — it can never touch the real ~/.crow; see scripts/run-suite.mjs). Node built-in test runner, no third-party framework. Single file: npm test -- tests/<file>.test.js (or raw node --test tests/<file>.test.js — but NOT from a systemd-launched shell without the scratch env). Also verify a server starts cleanly: node servers/<name>/index.js (ctrl-C to exit) or node servers/gateway/index.js --no-auth for the gateway.
CI red blocks ALL merges — including the operator's (decision Q6, 2026-07-18). The Tests workflow (.github/workflows/test.yml) runs the full suite + check-ports + build-registry --check + npm audit (critical/prod tier) on every PR and push to main, with no path filters — every commit gets check-runs now; an empty check-runs result on a current sha means something is WRONG, not normal. Query https://api.github.com/repos/<owner>/<repo>/commits/<sha>/check-runs (public, no auth) and confirm every run is completed/success before merging (the legacy commit-status API omits Actions and can read green when a check is red). Job keys suite/static-checks/audit are branch-protection contexts — never add a name: to those jobs or rename a key without updating protection. CI green does NOT cover schema migrations — SCHEMA_GENERATION bumps still require the manual migration rail (scripts/schema-migration-dryrun.sh). Notably any new bundle host port must be added to docs/developers/port-allocation.md or check-ports (scripts/check-port-allocation.js) fails. (See ~/.claude/CLAUDE.md Learnings 2026-06-28.)

Read the full file on GitHub · 63 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

2d ago First seen · 63 lines · 2,090 tokens per session scan A 416ce7fbfa45

Subscribe to this mod's changes

crow CLAUDE.md is an instructions file published in the GitHub repository kh0pper/crow (22 stars, last pushed 10d ago), licensed MIT. It adds 2,090 tokens to every session, about $0.0104 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Data and AI LLM Testing agent-framework ai-agents bot-builder home-server mcp

crow CLAUDE.md

Getting it into your agent

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

CLAUDE.md

Working with this repo

What this file has done since we first saw it

Other instructions, from other repositories

waku-agent CLAUDE.md

kungfu copilot-instructions.md

aeon CLAUDE.md

CrewClaw AGENTS.md

Agent-Worlds CLAUDE.md

nanobot CLAUDE.md