craft CLAUDE.md

A set of project instructions for the Craft plugin, covering its commands, skills, tests, documentation, and Git workflow. It defines how work moves between development, feature, and protected release branches.

In plain words
What is it for?
Use it when developing or releasing Craft, running its test suite, following branch rules, restoring work state, or using its commands and skills.
Why use it?
It reduces mistakes such as coding on the wrong branch, committing directly to the protected main branch, or declaring work complete without the required tests and release steps.

Instructions file

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/data-wise/craft/claude-md
Clone the repo
git clone --depth 1 https://github.com/Data-Wise/craft
Per session 1,233 This file is loaded in full into every session.
When invoked 1,233 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.

ModelPer sessionOnce invoked
Fable 5 $0.01233 $0.01233
Opus 5 $0.00616 $0.00616
Sonnet 5 $0.00247 $0.00247
Haiku 4.5 $0.00123 $0.00123

Measured 2d ago against content hash 7a2b553d6ea6, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

craft 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 · 103 lines

How it starts

The opening of the file, as written. The whole thing — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md - Craft Plugin

Active Work

48 commands · 41 skills · 2 agents · Docs · GitHub

/craft:finish — renamed from /craft:done (ADR-006, v4.2.0, breaking) · orchestrate:drive — spec-driven autonomous /goal loop · prompt-refiner skill + --refine flag (9 commands) · /craft:restore — combined git+.STATUS recap

Current Version: v4.6.0 | Tests: full suite via python3 -m pytest tests/ · tiers via /craft:test <unit|e2e|dogfood>

For project details, see plugin.json description and scripts/validate-counts.sh

Git Workflow

main (protected) ← PR only, never direct commits
  ↑
dev (integration) ← Plan here, branch from here
  ↑
feature/* (worktrees) ← All implementation work

Constraints

  • CRITICAL: Always start work from dev branch (git checkout dev)
  • Never commit directly to main
  • Never write feature code on dev
  • Always verify branch: git branch --show-current

Branch Protection

Branch Code Files .md Files Git Operations
main BLOCKED BLOCKED Commit/push BLOCKED
dev New: BLOCKED, Existing: allowed ALLOWED Commit/push allowed
feature/* ALLOWED ALLOWED All allowed

Override local hook: ask "unprotect dev" / "bypass branch guard" (dev/git skill, Operation 10 — session-scoped, auto-expires).

Quick Commands

Task Command
Run unit tests python3 tests/test_craft_plugin.py or /craft:test unit
E2E tests python3 -m pytest tests/test_plugin_e2e.py -v
Dogfood tests python3 -m pytest tests/test_plugin_dogfood.py -v
Integration tests python3 tests/test_integration_*.py
Dependency tests bash tests/test_dependency_management.sh
Validate counts ./scripts/validate-counts.sh
Pre-release check ./scripts/pre-release-check.sh <version>
Post-release sweep ./scripts/post-release-sweep.sh or --fix
Docs staleness check ./scripts/docs-staleness-check.sh or --fix
Build docs mkdocs build
Apply GitHub-side protection ask "apply GitHub branch protection [--repo OWNER/REPO]" (dev/git skill, Operation 9)
Smart routing /craft:do <task>
Pre-flight check /craft:check
Lint code /craft:code:lint
Release pipeline /release or /release -n
Brainstorm /craft:brainstorm
Orchestrate /craft:orch

Read the full file on GitHub · 103 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.

  1. 2d ago First seen · 103 lines · 1,233 tokens per session scan A 7a2b553d6ea6

Subscribe to this mod's changes

craft CLAUDE.md is an instructions file published in the GitHub repository Data-Wise/craft (4 stars, last pushed 16d ago), licensed MIT. It adds 1,233 tokens to every session, about $0.0062 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 2026-08-31.

Related

Other instructions, from other repositories