Multi-agent PRD framework for Claude Code — structured requirements creation and review using BABOK, perspective-based reading, and smell detection
vshidlovsky/prd-agents-framework
Agent
Reviews PRDs for completeness, API accuracy, and implementation readiness. Orchestrates parallel sub-reviewers with a three-phase matrix-driven process — scaffold, fill (parallel agents including a dedicated smell pass), verify — to guarantee full coverage. Use after a PRD draft exists and before approval.
vshidlovsky/prd-agents-framework
Agent
Judges a technical PRD review, decides the product questions it raises, and turns the survivors into revision tickets for the writer. Runs after prd-reviewer and before any revision cycle. Use when a review exists and someone must decide what is actually worth fixing.
vshidlovsky/prd-agents-framework
Agent
Canonical list of anti-patterns in requirements. Used by the PRD reviewer and its sub-agents when evaluating FR and AC quality.
vshidlovsky/prd-agents-framework
Agent
Drafts structured PRDs from plain-language initiative descriptions. Researches codebase and API docs, asks clarifying questions, writes a complete spec. Use when someone needs to create a new initiative spec.
vshidlovsky/prd-agents-framework
Agent
Detects project type, scans the repo, and drafts .claude/project-context.md automatically. Run once when adopting the PRD agents framework into a new project.
vshidlovsky/prd-agents-framework
Agent
Researches a codebase for a specific initiative. Traces code paths, extracts API contracts, documents business logic. Produces a structured research document. Use at the start of the PRD workflow.