Skip to main content
ClaudeWave
Skill58 repo starsupdated 2mo ago

fast-review

Mandatory post-delegation gate: checks output completeness, verifies acceptance criteria compliance, flags regressions, produces PASS/FAIL verdict. Use when checking delegated work against acceptance criteria, running post-delegation gate, validating agent output before acceptance, verifying sub-agent completed its assignment, or running post-delegation QA check.

Install in Claude Code
Copy
git clone --depth 1 https://github.com/monkilabs/opencastle /tmp/fast-review && cp -r /tmp/fast-review/src/orchestrator/skills/fast-review ~/.claude/skills/fast-review
Then start a new Claude Code session; the skill loads automatically.

SKILL.md

# Skill: Fast Review

## Contract

| Rule | Detail |
|------|--------|
| Trigger | After **every** delegation — no exceptions |
| Reviewer | Single sub-agent; Economy tier (Standard for premium/security work) |
| Verdict | PASS or FAIL with structured feedback |
| Retry | ≤2 retries on FAIL; 3rd FAIL → panel review |
 

## Procedure

### 1 — Collect Context

Issue + acceptance criteria, file diff, file partition, deterministic results (lint/test/build), agent self-report.

### 2 — Spawn Reviewer

Single `runSubagent`. Context = acceptance criteria, diff, partition, deterministic results **only** — no session history, no delegation prompt.

```js
runSubagent({ agentName: 'Reviewer', prompt: `Review against ACs:\n${criteria}\nDiff:\n${diff}\nGates: lint ✅ test ✅ build ✅` });
```

### 3 — Parse Verdict

```
VERDICT: PASS | FAIL
ISSUES:
- [severity:critical|major|minor] Description
FEEDBACK: Actionable feedback.
CONFIDENCE: low | medium | high
```

- **PASS** — no critical/major issues (minor noted, non-blocking).
- **FAIL** — any critical/major issue, or output format mismatch.

**Auto-PASS** (skip reviewer): pure research/no code changes; docs-only `.md` changes; ≤10 lines across ≤2 non-sensitive files with all deterministic gates passing.

> **Sensitive override:** Auth/middleware, DB migrations, RLS policies, security headers, CSP, env var schemas, CI/CD config always require review — even 1-line changes.

### 4 — Handle Verdict

| Outcome | Action |
|---------|--------|
| PASS | Log review; continue |
| FAIL 1–2 | Log; re-delegate same agent: "Retry N/2 — address listed issues" |
| FAIL 3 | Log `escalated: true`; load **panel-majority-vote** skill |
| Panel BLOCK ×3 | Dispute in `.opencastle/DISPUTES.md` (see **team-lead-reference** § Dispute Protocol) |

## Reviewer Prompt Template

See [REFERENCE.md](REFERENCE.md) for full reviewer prompt template.

## Logging

> **⛔ HARD GATE — Log the review before proceeding.**

```sh
npx opencastle log review --skill <name> --outcome pass|fail --reviewer "Reviewer" --mechanism sub-agent
```

## Integration & Overnight Mode

`on-post-delegate` Gate 5 (after deterministic Gates 1–4), ~5–15% token overhead. Overnight: upgrade one tier, escalate after 2 FAILs, checkpoint before panel.

## Anti-Patterns

- **Skipping fast review** — never, including "trivial" changes.
- **Panel as fast review** — wastes ~3× tokens.
- **Reviewer sees delegation prompt** — evaluate against acceptance criteria only.
- **Ignoring minor issues** — track; 3+ recurrences → ticket.
- **Force-accepting FAIL** — retry or escalate.
- **Skipping deterministic checks** — does NOT replace lint/test/build.
astro-frameworkSkill

Creates pages/layouts, defines content collections, configures hydration directives, and wires integrations. Use when adding or modifying Astro pages, layouts, components, or content collections. Trigger terms: Astro, content collection, client:load, client:visible, astro:content

browser-testingSkill

Drive real browsers via Chrome DevTools MCP: navigate pages, capture snapshots, run responsive checks, and collect console/perf traces. Use when the user mentions: 'validate UI change in Chrome', 'capture a screenshot', 'run responsive checks', or 'collect console logs'. Trigger terms: browser testing, DevTools, console logs, screenshot, responsive testing

cloudflare-platformSkill

Creates and deploys Cloudflare Workers, configures wrangler.toml bindings, sets up KV/D1/R2 storage and Durable Objects, manages Pages deployments, and implements edge function patterns. Use when building or deploying Cloudflare Workers, setting up Pages, working with KV/D1/R2 storage, configuring wrangler.toml, or deploying edge applications.

contentful-cmsSkill

Creates Contentful content types, queries entries via GraphQL/REST, runs CLI migrations, and manages assets and locales. Use when building or modifying Contentful content models, writing queries, or migrating content.

convex-databaseSkill

Convex reactive database patterns, schema design, real-time queries, mutations, actions, authentication, migrations, performance optimization, and component creation. Use when designing Convex schemas, writing queries/mutations, managing the Convex backend, setting up auth, migrating data, optimizing performance, or building Convex components.

coolify-deploymentSkill

Deploys applications, databases, and services on self-hosted Coolify instances, manages environments and env vars, runs infrastructure diagnostics, and performs batch operations. Use when deploying apps to Coolify, managing Coolify servers, debugging deployment issues, setting up databases, or managing Coolify infrastructure.

cypress-testingSkill

Writes Cypress E2E/component tests, configures `cy.intercept()` and `cy.session()`, authors custom commands, and wires CI artifacts. Use when creating E2E specs, component tests, or CI test pipelines. Trigger terms: cypress, e2e, component test, cy.intercept, cy.session

drizzle-ormSkill

Drizzle ORM schema definition, type-safe queries, relational queries, CRUD operations, transactions, migrations with drizzle-kit, and database setup for PostgreSQL, MySQL, and SQLite. Use when defining database schemas, writing queries or joins, managing migrations, setting up a new Drizzle project, or working with drizzle-kit.