Does it adapt its feedback based on the project stage?

Yes, it identifies the current stage (e.g., Planning Review vs. Launch Readiness) to focus on relevant priorities like scope vs. compliance.

Can I review a specific subset of stakeholders?

Yes, you can use the --perspectives flag followed by specific agent codes like 'eng,design,exec' to limit the scope of the review.

Which perspectives are included in the PRD review?

The panel includes 7 agents: Engineering, Design, Executive, Legal, UX Research, a Skeptic (devil's advocate), and the Customer Voice.

How does it handle conflicting feedback between agents?

The skill synthesizes output by identifying 'Convergent feedback' where agents agree and 'Conflicting perspectives' that require specific PM judgment.

Where does the skill look for my PRD files?

The skill scans the 'thoughts/shared/pm/prds/' directory for active markdown files and provides a list if a specific name isn't provided.

PRD Review Panel

Name: PRD Review Panel
Author: coalesce-labs

bycoalesce-labs

•

协作工具

Conducts comprehensive, multi-perspective product requirement reviews using seven specialized AI sub-agents in parallel.

The PRD Review Panel streamlines the product development lifecycle by simulating a cross-functional review committee within Claude Code. It spawns seven distinct AI personas—Engineering, Design, Executive, Legal, UX Research, Skeptic, and Customer Voice—to analyze PRDs for technical feasibility, strategic alignment, and potential risks. This skill helps product managers and developers identify critical blockers, unvalidated assumptions, and edge cases before presenting to human stakeholders, ensuring higher quality documentation and smoother project execution.

主要功能

01Persona-based reviews including Engineering, Legal, and UX Research

02Contextual routing that scans local project directories for active PRDs

03Automated synthesis of feedback into blockers, gaps, and enhancements

04Lifecycle-aware analysis tailored to specific stages like Kickoff or Launch Readiness

059 GitHub stars

06Parallel multi-agent execution for rapid, 360-degree feedback

使用场景

01Simulating executive and legal pushback to prepare for high-stakes stakeholder meetings

02Stress-testing user experience flows and accessibility requirements during the design phase

03Pre-screening product requirements for technical complexity and hidden dependencies

主要功能

01Persona-based reviews including Engineering, Legal, and UX Research

02Contextual routing that scans local project directories for active PRDs

03Automated synthesis of feedback into blockers, gaps, and enhancements

04Lifecycle-aware analysis tailored to specific stages like Kickoff or Launch Readiness

059 GitHub stars

06Parallel multi-agent execution for rapid, 360-degree feedback

使用场景

01Simulating executive and legal pushback to prepare for high-stakes stakeholder meetings

02Stress-testing user experience flows and accessibility requirements during the design phase

03Pre-screening product requirements for technical complexity and hidden dependencies