一键导入
brainstorm-templates
Sidecar data for /brainstorm workflow — Brainstorm (open), Decision (finalized), and Research document templates loaded just-in-time.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Sidecar data for /brainstorm workflow — Brainstorm (open), Decision (finalized), and Research document templates loaded just-in-time.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Companion skill for the /open workflow. Automates workspace diagnostic steps when starting a new session on a project. Collects active session status, project contracts, active task metrics, compact memory presence, roadmap phases, pending brainstorms, and git status.
Sidecar data for /docs workflow — document templates loaded just-in-time when generating docs.
Modular HTML rendering engine for PARA Workspace. Converts Markdown documents, code graphs, and standalone files into interactive, self-contained HTML pages with consistent Notion-inspired theming. Triggers when the user asks to compile docs to HTML, render graph visualization, export markdown to HTML, view docs statically, or preview documents.
Sidecar data for /plan workflow — Detail Plan and Roadmap templates loaded just-in-time.
Sidecar data for /qa workflow — Red Team personas, dimension checklists, and QA report templates loaded just-in-time.
Sidecar data for /remote workflow — namespace conventions and naming patterns loaded just-in-time. Extensible by user.
| name | Brainstorm Templates |
| description | Sidecar data for /brainstorm workflow — Brainstorm (open), Decision (finalized), and Research document templates loaded just-in-time. |
| source | catalog |
Sidecar Skill for the
/brainstormworkflow. Contains document templates that the Agent loads only when saving structured output (Step 4).Pattern: Workflow = Logic → Sidecar Skill = Data Router. The
/brainstormworkflow instructs the Agent to read this skill at save time.
/brainstorm → Step 4 (Open brainstorm): load references/templates/brainstorm.md/brainstorm → Step 4 (Finalized decision - General): load references/templates/decision.md/brainstorm → Step 4 (Finalized decision - Architectural/ADR): load references/templates/adr.md/brainstorm → Step 4 (Research extraction): load references/templates/research.md (if user consents)| File | When | Purpose |
|---|---|---|
references/templates/brainstorm.md | Step 4 — Open exploration | Living document for ongoing research, appendable across sessions |
references/templates/decision.md | Step 4 — Finalized decision | Frozen document for concluded general brainstorms with chosen option |
references/templates/adr.md | Step 4 — Finalized ADR | Frozen document for concluded architectural brainstorms (ADR) |
references/templates/research.md | Step 4 — File 2 (user consent) | Document structure for extracted research |
Convention: Data files live in
references/(nottemplates/at skill root). This follows the Sidecar Skill convention formalized in v1.7.6.3.
| Type | Directory | Filename Pattern | Has Date Prefix |
|---|---|---|---|
| Open Brainstorm | artifacts/brainstorms/ | brainstorm-[topic-slug].md | ❌ No date — living document |
| Decision | artifacts/para-decisions/ | brainstorm-[YYYY-MM-DD]-[topic-slug].md | ✅ Yes — point-in-time snapshot |
brainstorms/brainstorm-[topic].md (Open, append across sessions)
│
▼ When user finalizes a decision
para-decisions/brainstorm-YYYY-MM-DD-[topic].md (Frozen, cross-refs brainstorm)
para-decisions/ — NOT move the brainstorm.brainstorms/ as historical record.When Step 4 triggers Research extraction:
When brainstorming about code architecture or implementation:
🔍 If project has
.beads/graph/, readpara-graph §3.3.3for codebase understanding pipeline (query structure → context bundle → informed discussion). If no graph → proceed with source-only brainstorming.