---
title: "calm-craft"
description: "Specs as the source of truth, a chunked delivery loop, and decided-and-enforced code conventions, as portable agent skills."
canonical: https://agentpluginsdirectory.com/plugins/calm-craft
last-updated: 2026-09-27
---

# calm-craft
Specs as the source of truth, a chunked delivery loop, and decided-and-enforced code conventions, as portable agent skills.
- Slug: calm-craft
- Publisher: Calm Tech Ltd
- Repository: https://github.com/calmtechltd/calm-craft
- Manifest: plugin.json
- Version: 0.3.1
- License: MIT
- Category (editorial): code-review
- Skills: 31 (ask-questions, author-implementation-plan, branch-cleanup, branch-self-review, bug-regression-red-green, clean-code-slop, coderabbit-review-implement-all, coderabbit-review-implement, coderabbit-review-triage, conventions-audit, conventions-decide, conventions-migrate, conventions-revisit, engineering-setup, ready-for-pr, run-implementation-plan-all, run-implementation-plan, spec-assess-coverage, spec-audit-drift, spec-author-from-impl, spec-author-greenfield, spec-author-tests, spec-gap-sweep, spec-harvest-discussion, spec-maintain-on-ship, spec-plan-gap, spec-storyboard-journey, spec-triage-bug-report, spec-visualize, update-pr, write-tests)
- MCP servers: 0
- Stars: 2
- Repository created: 2026-08-13
- Repository last pushed: 2026-09-25
- Publisher type: Organization
- Listing: https://agentpluginsdirectory.com/plugins/calm-craft
- Schema: https://agent-plugins.org/schemas/1.0.0/plugin.schema.json

## What calm-craft does, in the publisher's words

Free and open source, from Calm Tech Ltd. MIT licensed: use it, change it, ship it commercially. We want nothing for it.

An Agent Plugin providing three things that make coding agents produce work you can trust:

- Specs: the product's intent, written down in an addressable form, so an agent has a source of truth that doesn't move.
- A delivery loop: design → chunked plan → one chunk per pass, looped until the plan is done → gates → self-review.
- Conventions: decided once, enforced by lint where a machine can enforce them, ambient where it can't.

Portable by design: skills live in skills/ per the Agent Plugins v1 spec, so this works in Codex, Cursor, VS Code, Copilot, and anything else implementing the standard..claude-plugin/ sits alongside for Claude Code, which uses its own manifest location; both read the same skills.

From the project README, punctuation lightly normalized. Full text: https://raw.githubusercontent.com/calmtechltd/calm-craft/HEAD/README.md

## Skills

- ask-questions: Surface consequential open decisions in the current task and ask through the host's available question tools. Use when asked for questions or when necessary facts remain unresolved; reuse settled decisions and authorized defaults.
- author-implementation-plan: Turn scoped requirements into a dependency-ordered implementation plan with reviewable chunks, acceptance criteria, and exclusions. Plans only; preserves supplied decisions and selects proportionate verification.
- branch-cleanup: Survey local branches, worktrees, and Graphite stacks for cleanup, including squash-merged PRs. Use for branch or worktree cleanup and merge-status questions. Report merge evidence and ownership blockers before authorized local deletion; remote branches are report-only.
- branch-self-review: Review a branch or the current task's changes for functional bugs, permission and tenancy holes, and maintainability problems. Use when the user asks to review changes or an implementation workflow calls for final review. Reports findings without editing product code or running whole-repository gat…
- bug-regression-red-green: Fix a bug by reproducing it with a failing test, then fixing the code until green and preserving regression protection. Use when the user says "write a regression test", "red test first", "fix this bug properly", "make sure this never breaks again", or when fixing incorrect behaviour with no test c…
- clean-code-slop: Find and simplify low-value tests, redundant wrappers, duplicated business rules, speculative abstractions, and inconsistent code. Use only when the user explicitly requests a code-slop audit or cleanup, or invokes clean-code-slop. Honour review-only requests; otherwise make verified, behaviour-pre…
- coderabbit-review-implement-all: Publish CodeRabbit review fixes and resolve the review. Runs coderabbit-review-implement first if the local fixes are not done, commits and publishes them to the current PR, resolves inline threads with GraphQL, and when necessary posts one PR-level @coderabbitai resolve summary for completed revie…
- coderabbit-review-implement: Apply verified CodeRabbit triage fixes locally. Use when the user requests local review fixes; does not commit, publish, reply, or resolve unless a separate operation is explicitly authorized.
- coderabbit-review-triage: Download CodeRabbit PR review feedback, save raw comments under.active/, verify findings against the codebase, and produce a triage breakdown of obvious fixes, skips, and items needing user input. Fetch via GitHub GraphQL review threads: never GitHub MCP addComment, never gh pr comment, never RES…
- conventions-audit: Review a diff or file scope against recorded conventions, reporting verified violations and enforcement gaps. Reports only; avoids duplicating existing compiler, formatter, lint, or CI checks.
- conventions-decide: Record coding convention decisions and implement their agreed enforcement using existing repository tools. Use for choosing or revising standards; preserve settled decisions and distinguish chosen answers from defaults.
- conventions-migrate: Apply an agreed convention change to existing code in reviewable batches and tighten its enforcement. Use for convention migrations; select checks for the transformation instead of running tests after every batch.
- conventions-revisit
- engineering-setup
- ready-for-pr
- run-implementation-plan-all
- run-implementation-plan
- spec-assess-coverage
- spec-audit-drift
- spec-author-from-impl
- spec-author-greenfield
- spec-author-tests
- spec-gap-sweep
- spec-harvest-discussion
- spec-maintain-on-ship
- spec-plan-gap
- spec-storyboard-journey
- spec-triage-bug-report
- spec-visualize
- update-pr
- write-tests

Descriptions come from the frontmatter of each SKILL.md, punctuation lightly normalized.
