PM Critic — Development agent for Claude Code
Self-checks a product brief for vague metrics, missing evidence, and fuzzy scope before /speccraft:pm:review.
How to install PM Critic
Installs to ~/.claude/agents/dcstolf-speccraft-pm-critic.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/DCSTOLF/speccraft/HEAD/agents/pm-critic.md -o ~/.claude/agents/dcstolf-speccraft-pm-critic.md Restart Claude Code, or start a new session, for it to be picked up.
What PM Critic does
name: pm-critic description: "Self-checks a product brief for vague metrics, missing evidence, and fuzzy scope before /speccraft:pm:review. Single-model self-check, not a substitute for cross-review." tools: [Read] model: opus
You are the pm-critic. You run a cheap, single-model self-check on a draft `brief.md` BEFORE the cross-model `cross-reviewer` pass in `/speccraft:pm:review`. You are a lightweight self-check, not a substitute for the cross-model review, and you do not approve or b
Alternatives in Development
- Startup Metrics Framework — Comprehensive guide to tracking, calculating, and optimizing key performance metrics for different s... - wsho 31.9k ★
- Cross Linker — Scan recent or specified notes for missing wikilinks 4.6k ★
- Self Critic 3.7k ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Integration Critic
Adversarially reviews the COMBINED (union) tree of a multi-node ATLAS-WEAVE run through the SEAM lens — the cr
PM Author
Drafts and refines a product brief.md via Socratic interviewing. Use during /speccraft:pm:new.
Sylvia
Product-scope agent for fuzzy requests, backlog choices, and deciding the smallest useful thing to build. Defi
Cursor
Cursor slot (claude substitute) — implementation / cross-check via Cursor CLI
Kimi
Kimi slot (claude substitute) — implementation / cross-check via Kimi Code CLI
Story Analyzer
Use when analyzing parsed stories for discrepancies, ambiguities, contradictions, and coverage gaps. Surfaces
Related Skills
Claude Plugin
Personal skills for Codex and Claude Code, invoked as superb:. craft turns a vague product idea into a decisio
Incident Triage Agent
An AI-powered agent built with AWS Bedrock, Claude, and Strands Agents SDK that monitors system alerts, connec
Nexpath
Local-first AI coding workflow for vibe coders, indie hackers, technical founders and product managers — catch