prd-review-panellisted

## Why this exists, and what I'd change **Why it exists.** A single-perspective review misses the angle you're weakest on. Seven sub-agents (engineer, designer, exec, legal, UXR, skeptic, customer voice) catch what one reviewer misses. The whole loop runs in about 90 seconds. **Design tradeoffs.** - **Seven fixed personas instead of dynamic reviewers.** Each agent has a defined lens and a defined checklist. Cost: if your PRD really needs a "data scientist" or "compliance" view, you add a new sub-agent file before running this. - **Output is seven separate critiques plus a synthesis.** Not one merged review. Cost: nine paragraphs to read instead of one. Readers skim the synthesis and miss specific findings. - **Severity isn't ranked across agents.** Every critique is equal weight in the output. Cost: a tiny UX nit and a fatal legal flaw look the same on the page. **What I'd change.** Auto-rank the seven critiques by severity so the top three issues across all reviewers surface first. You can stop reading after that if time is short. --- ## Purpose Get comprehensive feedback on your PRD from 7 different perspectives in parallel: Engineering, Design, Executive, Legal, UX Research, Skeptic, and Customer Voice. Catches gaps, challenges assumptions, and surfaces conflicts before stakeholder review. ## Usage - `/prd-review-panel` - Review a PRD with all 7 sub-agents - `/prd-review-panel [prd-name]` - Review specific PRD - `/prd-review-panel --perspectives "eng,design,exec"`