prd

Generates a development-ready Product Requirement Document from a BRD using parallel agent teams.

4|Updated Mar 14, 2026
One-click install
npx skills add https://github.com/VoxTechnologies/transmute-framework --skill prd-voxtechnologies
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: prd
Source: https://github.com/VoxTechnologies/transmute-framework/tree/main/skills/prd
Command: npx skills add https://github.com/VoxTechnologies/transmute-framework --skill prd-voxtechnologies

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Translating a Business Requirement Document into a complete, engineering-ready PRD is slow and error-prone, often producing gaps in traceability, missing error states, and inconsistent specifications across features. ## Core Features & Use Cases - Multi-Agent PRD Generation: Spawns five specialized agent teammates covering product strategy, user stories, screen specifications, API and technical architecture, and quality/operations to produce 18 numbered PRD files. - Full Traceability and Coverage Audits: Builds a Feature Decomposition Map linking every BRD requirement (BR/FR/NFR) to epics, user stories, screens, and API endpoints, with structural consistency and dependency-cycle validation. - Quality Review and Remediation: Runs completeness, consistency, and technical review agents, then remediates CRITICAL and HIGH findings with a logged review trail. - Use Case: After completing Stage 1 (BRD) of the Transmute pipeline, run Stage 2 to produce a full PRD in ./plancasting/prd/ with user stories, screen specs, API schemas, and a testing strategy ready for Stage 3 scaffolding. ## Quick Start Ask the assistant to generate the PRD from the BRD by running Stage 2 of the transmute pipeline.

Frequently Asked Questions about prd

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a PRD from a BRD automatically?

Run Stage 2 of the Transmute pipeline after completing Stage 1. The skill reads all files in ./plancasting/brd/ and tech-stack.md, then spawns five agent teammates to produce 18 PRD files covering stories, screens, APIs, and testing.

What inputs are required before running PRD generation?

Stage 2 requires a populated ./plancasting/brd/ directory from Stage 1 and a tech-stack.md file from Stage 0 containing a Session Language setting. If either is missing, the skill stops with an instruction to run the earlier stage first.

What output files does the PRD generation produce?

It produces 18 numbered markdown files in ./plancasting/prd/ covering product overview, feature map, release plan, epics, user stories, user flows, screen specs, architecture, data model, API specs, testing, and risks, plus README.md, _context.md, and _review-log.md.

Does the PRD skill support BaaS backends like Convex or Firebase?

Yes. For function-based backends such as Convex or Firebase, API specifications document function signatures, argument schemas, return types, and error codes instead of HTTP methods, paths, and status codes.

What happens if the BRD has gaps or contradictions during PRD generation?

Issues are classified as BLOCKING, CRITICAL-BUT-RECOVERABLE, or NON-BLOCKING and documented in _brd-issues.md. Generation continues with marked assumptions where possible; only empty or corrupted BRD files cause a stop. Stage 2B resolves the issues afterward.

Can the PRD use feature flags for phased rollout?

No. The skill enforces a full-build approach where all features ship in a single release. Feature flags are permitted only as operational kill switches, A/B experiment flags, or permission gating, never for phased rollout.