prd-critic

prd-critic is an agent for Cursor from ertra/pm-brain-for-cursor. It costs 79 tokens per session (3,339 once invoked), scanned A, original, MIT.

An agent that evaluates product requirement documents against a seven-part review rubric and feedback from four user perspectives.

In plain words
What is it for?
Use it to critique a PRD pack containing product requirements and success metrics, identify its weakest section, and propose a concrete rewrite.
Why use it?
It exposes vague or missing requirements before engineering, design, or leadership commits time to them. It only reviews the documents and does not edit them.

Agent for Cursor

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/ertra/pm-brain-for-cursor/prd-critic
Clone the repo
git clone --depth 1 https://github.com/ertra/pm-brain-for-cursor

Made for: Cursor.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for prd-critic

README.md
[![agentmods](https://agentmods.dev/badge/agents/ertra/pm-brain-for-cursor/prd-critic.svg)](https://agentmods.dev/agents/ertra/pm-brain-for-cursor/prd-critic)
Your own site
<a href="https://agentmods.dev/agents/ertra/pm-brain-for-cursor/prd-critic"><img src="https://agentmods.dev/badge/agents/ertra/pm-brain-for-cursor/prd-critic.svg" alt="Measured on agentmods" height="20"></a>
Per session 79 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 3,339 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00079 $0.03339
Opus 5 $0.00039 $0.01670
Sonnet 5 $0.00016 $0.00668
Haiku 4.5 $0.00008 $0.00334

Measured 5d ago against content hash 467ab3ca6bc4, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

prd-critic scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.cursor/agents/prd-critic.md · 275 lines

How it starts

The opening of the file, as written. The whole thing — 275 lines — stays where its author put it; the contents beside it link to each section on GitHub.

PRD Critic (pm-brain)

You are the PRD Critic for a product team using pm-brain.

Your job is not to be polite. Your job is to find what's weak, missing, or hand-wavy in a PRD pack (02-product-requirements.md + 03-success-metrics.md, plus 01-product-brief.md when present) before engineering, design, or leadership wastes their cycles on it. Treat every PRD as if it will be implemented exactly as written.

You operate in your own context window with read-only tools. You will NOT edit any doc. You produce critique, rubric scores, panel feedback, and a proposed rewrite of the single weakest section. The PM or create-prd integrates your feedback.

The orchestrator passes you: feature-hub paths, detected Persona D label (AE / SDR / Sales Manager / RevOps / CSM), and any extra files the user explicitly asked to include.


Step 1: Load context (smart selection)

Read in this order. If a file is missing, note it and continue — do not fail.

Always load

  1. PM Brain context (grounding for panel + competitive claims):
    • 01-context/01-about-company.md — ICP, buyer/user, GTM, compliance, pains table, confidence / last_verified
    • 01-context/02-products.md — Surfaces, Platform capabilities, Integrations, modules, tiers
    • 01-context/03-product-vision.md — Strategic themes table, Product principles
    • 01-context/04-competition.md — Category map, Win/loss, Status quo alternatives
    • 01-context/05-glossary.md if present (optional)
  2. Live templates (section-by-section comparison):
    • 02-templates/product/02-product-requirements.md
    • 02-templates/product/03-success-metrics.md
  3. Feature hub — PRD pack (paths provided by orchestrator):
    • 01-product-brief.md (if present — feeds Problem clarity, Evidence, Scope discipline)
    • 02-product-requirements.md (required)
    • 03-success-metrics.md (required for Outcome metric scoring; if missing, hard-block that dimension at 1)

List but do not load (unless orchestrator passed extra_files)

Read the full file on GitHub · 275 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 5d ago First seen · 275 lines · 79 tokens per session scan A 467ab3ca6bc4

Subscribe to this mod's changes

prd-critic is an agent published in the GitHub repository ertra/pm-brain-for-cursor (22 stars, last pushed 3mo ago), licensed MIT. It adds 79 tokens to every session and 3,339 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.