Add FeatureBrainstormAgent for structured ideation and feature candidate prioritization

[codegen-sh]

Overview

Implements FeatureBrainstormAgent to fill a critical gap in the DevAgent workflow by facilitating divergent ideation before convergent research and specification. This agent helps teams explore solution spaces, generate feature ideas, and prioritize opportunities before committing to detailed investigation.

What's New

Agent Brief: FeatureBrainstormAgent.md

• Mission: Structured ideation sessions that generate, cluster, and prioritize feature candidates aligned with product mission
• Three operating modes:
  • Exploratory: Open-ended ideation from problem statements (15-30 ideas → clustering → top 3-5 candidates)
  • Targeted: Solutions for specific mission gaps with constraints
  • Expansion: Variations and enhancements on existing feature concepts
• Integration: Positioned between ProductMissionPartner and ResearchAgent in workflow
• Outputs: Research-ready packets with prioritized candidates, evaluation matrices, and formulated research questions

Template: brainstorm-packet-template.md

• Structured format for brainstorm outputs including:
  • Problem statement and mode selection
  • Full idea list (divergent phase)
  • Clustered themes with pattern identification
  • Evaluation matrix (mission alignment, user impact, feasibility, effort)
  • Prioritized top 3-5 candidates with detailed rationale
  • Research questions for ResearchAgent
  • Parking lot for future ideas
  • Session metadata and decision log

Workflow Integration

• Updated AGENTS.md to register agent in roster
• Storage strategy: .devagent/product/brainstorms/YYYY-MM-DD_<topic>.md
• Handoff protocol: Outputs feed ResearchAgent with prioritized candidates
• Traceability: Updates decision journal and guiding questions

Workflow Position

ProductMissionPartner (mission updates)
  → [NEW] FeatureBrainstormAgent (generate & prioritize feature candidates)
    → ResearchAgent (evidence gathering for TOP candidates)
      → SpecArchitect (spec drafting)
        → TaskPlanner (implementation planning)
          → TaskExecutor (code)

Key Design Decisions

1. Mission-grounded evaluation: Every idea scored against established mission metrics and constitution principles
2. Research-ready outputs: Brainstorm results include formulated research questions for seamless handoff
3. Lightweight governance: Executing developer has standing approval to run sessions (no approval gates)
4. Multi-technique generation: Prompt-based, constraint-based, analogy, SCAMPER, "How Might We" framing
5. Convergent deliverables: Prioritized candidates (not just divergent idea lists) for actionable next steps

Research Foundation

Based on comprehensive research packet: .devagent/research/2025-10-01_feature-brainstorm-agent.md

Industry patterns analyzed:

• Google Mixboard AI-powered ideation
• Miro AI clustering and collaboration
• Multi-agent brainstorming techniques
• Design thinking convergent/divergent frameworks

Files Changed

• .devagent/agents/FeatureBrainstormAgent.md — Complete agent brief following template structure
• .devagent/templates/brainstorm-packet-template.md — Reusable template for consistent outputs
• AGENTS.md — Registered new agent in roster and updated workflow sequence
• .devagent/product/brainstorms/ — New directory for brainstorm artifact storage

Acceptance Criteria Met

✅ Agent brief follows agent-brief-template.md structure
✅ All research findings incorporated (workflow position, capabilities, modes, outputs)
✅ Mission statement defines goal, boundaries, and success signals
✅ Three operating modes documented with specific workflows
✅ Integration points with ProductMissionPartner and ResearchAgent specified
✅ Storage strategy defined for brainstorm artifacts
✅ Workflow aligns with DevAgent patterns (standing approval, traceability)
✅ Brainstorm packet template created with structured sections
✅ Follow-up hooks identify downstream agents and handoff protocols
✅ Failure & escalation patterns documented

Next Steps

1. Try running a brainstorm session using the new agent
2. Validate outputs meet ResearchAgent input expectations
3. Gather feedback on template structure and evaluation criteria
4. Consider future enhancements (external input integration, custom scoring frameworks)

---

Requested by: Jake Ruesink
Research by: DevAgent Team
Implementation: Codegen

---

💻 View my work • 👤 Initiated by @jaruesink • About Codegen
⛔ Remove Codegen from PR • 🚫 Ban action checks

[coderabbitai]

Important

Review skipped

Draft detected.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

✨ Finishing touches

🧪 Generate unit tests

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch feature/brainstorm-agent

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}