brainstorm-path

Skill: Brainstorm Path

Manage brainstorm decision paths — recording rounds, querying history, creating branches, evaluating thresholds, and rendering visualizations.

Overview

The brainstorm path system records the evolution of innovation candidates across brainstorming rounds. It enables:

• Round-by-round recording with scores and decisions
• Threshold-based pass/iterate decisions
• Branching from any historical node
• Restoring abandoned innovations
• Visualizing the complete decision path

All operations are performed via the oh-my-patent CLI, which wraps the TypeScript APIs. Do not manually write JSON files — use the CLI commands below.

CLI Access

node <oh-my-patent-dir>/dist/cli.js path <subcommand> [options]

Where <oh-my-patent-dir> is the path to the oh-my-patent plugin directory (e.g., ./oh-my-patent from the workspace root).

Subcommands

Initialize a new path

node dist/cli.js path init <project-path>

Creates the .brainstorm/ directory structure with path.json, nodes/, and snapshots/.

Output: { "ok": true, "pathId": "path-..." }

Record a brainstorm round

node dist/cli.js path record <project-path> \
  --round <N> \
  --data '<json-string>'

Or read from a file:

node dist/cli.js path record <project-path> \
  --round <N> \
  --data @/path/to/round-data.json

The --data JSON structure:

{
  "projectId": "05-crypto-hsm-cache-concurrency",
  "topic": "加密机缓存并发优化",
  "agentOutputs": [
    {
      "agentId": "patent-innovation-architect",
      "outputFile": "references/brainstorm_round1_innovation_architect.md",
      "summary": "生成3个候选方案",
      "keyPoints": ["A1: 三级缓存架构", "A2: 失效协议"]
    }
  ],
  "innovations": [
    {
      "id": "INN-001",
      "title": "安全策略约束的多级缓存并发控制系统",
      "problem": "HSM高并发场景重复计算、缓存风暴",
      "coreSolution": ["三级缓存架构", "单flight并发控制"],
      "differences": ["vs R6/R7: HSM场景特化"],
      "status": "active"
    }
  ],
  "scores": [
    {
      "innovationId": "INN-001",
      "novelty": 8,
      "creativity": 8,
      "practicality": 9,
      "businessValue": 7,
      "weightedScore": 8.1
    }
  ],
  "decision": {
    "action": "ITERATE",
    "reason": "综合分 8.0 < 阈值 8.5",
    "recommendations": ["强化缓存键策略绑定的差异性描述"]
  }
}