spec-review

Three LLMs review your design plan in parallel — paste 30 lines into your CLAUDE.md, no install needed.

Agent 说'明白了'——你的请求它真的看懂了吗？FindGap 在你按 Enter 前找出会导致返工的 gap。Trust Chain 第一节点。

Local-first roadmap and spec-review workbench for humans and agents. Repo-local roadmap files plus per-file spec sessions for line-anchored review and suggestions.

Structured multi-agent review for plans, specs, and code with iterative debate, patch-based auto-apply, and a full audit trail.

3-round cross-agent spec/plan review pipeline. Reviewer audits findings and author settles them across fresh sessions, with a strict final round.

Paste a draft PRD and get the questions engineering will send back, ranked by whether the gap blocks the build, sign-off or launch. 22 deterministic rules, every one printed in the app. No model, no API key, runs entirely in the browser.

Claude Code skill (also Codex CLI): audit a spec or design doc before you build it. 3 to 5 independent AI agents review in parallel (completeness, premortem, cut scope) and return a change plan; the spec itself is never edited. spec 定稿前的完整性稽核＋多面向 premortem

AI-native decision gate for PRD/spec markdown. Reads your plan, finds fragile paths, and blocks them before they become code.

Read-only pre-merge spec review, delivered as a Claude Code skill. Compares a feature branch's git diff against its Notion spec page (via the Notion MCP plugin), evaluates AC quality, page conventions, and spec drift, and produces a forwardable written report.

Specialized Claude Code skills for spec review, design fidelity, feature completeness, and multi-layer testing. Complements the superpowers plugin.

Small, opinionated Agent Skills for bounded, decision-oriented workflows.