Keystone Review

Run the four review subagents on the current diff (functional, security, risk, deployment). Triggers: keystone review; keystone:review; /keystone:review; review the diff; run review subagents; semantic review.

44 stars 3 forksAI, LLMs & Agents

Install

skills CLI (Claude Code, Cursor, Codex and more)
npx skills add tacoda/keystone --skill keystone-review
Manual install
git clone --depth 1 https://github.com/tacoda/keystone.git /tmp/keystone
mkdir -p ~/.claude/skills
cp -r /tmp/keystone/.agents/skills/keystone-review ~/.claude/skills/keystone-review

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# keystone:review — semantic review of the current diff

Phase 5 of the six-phase lifecycle. Fires the four review
subagents against the in-progress diff:

- **review-functional** — does the change implement the spec?
- **review-security** — auth, input validation, secrets, injection.
- **review-risk** — blast radius, data integrity, observability.
- **review-deployment** — migrations, rollback, config, infra deps.

Each subagent returns a structured finding list. Semantic check — runs
after computational hooks (lint / type / test / build) pass.

Full
discipline at `.agents/rules/process/review.md`.

## Run

**Run the four review subagents on the current diff.** Semantic check, post-verify. Read [`.agents/rules/process/review.md`](../../rules/process/review.md).

## Activities

Run the four review subagents. If the agent supports parallel subagents (Claude Code, etc.), spawn them concurrently: **all four in a single message**. Then combine their findings. Agents without subagent parallelism run them sequentially.

| Subagent | Doc |
|---|---|
| review-functional | [`.agents/agents/review-functional.md`](../../agents/review-functional.md) |
| review-security | [`.agents/agents/review-security.md`](../../agents/review-security.md) |
| review-risk | [`.agents/agents/review-risk.md`](../../agents/review-risk.md) |
| review-deployment | [`.agents/agents/review-deployment.md`](../../agents/review-deployment.md) |
Read the full skill on GitHub

Details

Skill name
keystone-review
Path
.agents/skills/keystone-review/SKILL.md
License
MIT
Repo updated
17 November 2026

FAQ

What does the Keystone Review skill do?

Run the four review subagents on the current diff (functional, security, risk, deployment). Triggers: keystone review; keystone:review; /keystone:review; review the diff; run review subagents; semantic review.

How do I install the Keystone Review skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add tacoda/keystone --skill keystone-review. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Keystone Review popular?

It lives in tacoda/keystone, which has 44 GitHub stars and 3 forks.

Related skills

Keystone Allocate

44

tacoda/keystone

Move an existing agent setup into the charter and cut always-loaded context: run keystone allocate, review its classifications, then apply. Triggers: keystone allocate; keystone:allocate; /keystone:allocate; convert my .claude setup; migrate .cursorrules; trim CLAUDE.md; reduce context bloat.

AI, LLMs & Agents

Keystone Audit

44

tacoda/keystone

Full dual-flywheel audit — Learning (capture from review) + Pruning (remove dead rules). Periodic charter hygiene. Triggers: keystone audit; keystone:audit; /keystone:audit; run audit; audit the charter; dual flywheel audit.

Data & Analytics

Keystone Bootstrap

44

tacoda/keystone

One-time initial charter scaffold — detect stack, seed state, scaffold idioms, classify hooks, build the globs index. Triggers: keystone bootstrap; keystone:bootstrap; /keystone:bootstrap; run bootstrap; bootstrap the charter; bootstrap the keystone charter; first-time setup.

Data & Analytics

KrillClaw

57

javimosch/supercli

Use this skill when the user wants to use KrillClaw for The world's smallest AI agent runtime. 49KB. Written in Zig. Zero dependencies..

AI, LLMs & AgentsIn 1 list

Keystone New Adapter

44

tacoda/keystone

Define a host keystone has no built-in profile for (.agents/hosts/<name>.json): pick native, link, copy, or index per kind, then project to it. Triggers: keystone new adapter; keystone:new-adapter; /keystone:new-adapter; add a host; support a new coding agent; project to another agent.

AI, LLMs & Agents

OpenGlass

57

javimosch/supercli

Use this skill when the user wants to use OpenGlass for Turn any glasses into AI-powered smart glasses.

AI, LLMs & AgentsIn 1 list

Browse the most popular Claude skills, all ai, llms & agents skills, or write your own with the skills generator.