edisonlzy/divide-into-topic

Divide Into Topic

Framework Context

See Framework Structure for the canonical Obsidian knowledge tree structure definition.

Configuration

See Configuration Check Flow before executing this skill.

---

Subskill: Divide Into Topic

Goal

Divide a large topic into smaller, well-defined sub-topic nodes in the knowledge tree. This skill breaks down a broad topic into its core sub-topics through web research and user collaboration, making the knowledge tree more granular and navigable.

Input

• Target topic name or path (the topic to dive into)
• Vault root path (resolved from configuration obsidian_vault_path)

Execution Steps

Flow Overview

flowchart TD
    A[Start: identify target topic] --> B[Locate topic node via Retrieve Topic Skill]
    B --> C[Negotiate division direction with user]
    C --> D{Direction confirmed?}
    D -->|No| C
    D -->|Yes| E[Dispatch parallel web research agents]
    E --> F[Synthesize research results]
    F --> G[Propose sub-topics to user]
    G --> H{Sub-topics confirmed?}
    H -->|No / adjust| G
    H -->|Yes| I[Dispatch parallel Create Topic agents]
    I --> J[✅ Output created sub-topic tree]

---

Step 1: Locate Target Topic

Invoke the Retrieve Topic Node subskill using the user-specified topic name as the query.

• On success: obtain the confirmed topic directory path and read its README.md to understand the current scope and existing sub-topics.
• If not found: inform the user and offer to create the topic first using the Create Topic Skill.

---

Step 2: Negotiate Division Direction (Human-in-the-Loop)

Before researching, align with the user on how to divide the topic.

1. Initial proposal: Based on the topic's README and any general knowledge, propose 3–5 candidate research directions. For each, provide a one-sentence rationale.

   Example for topic "Frontend":

   I can divide "Frontend" into the following sub-areas:
   
   1. Core language fundamentals (HTML, CSS, JavaScript)
   2. Frontend frameworks & libraries (React, Vue, Angular)
   3. Engineering & toolchain (bundlers, CI/CD, testing)
   4. Performance optimization
   5. Career path & skill progression
   
   Which division angle(s) do you want to use? You can pick multiple or suggest your own.
   

2. Use web search to enrich your understanding of the topic before proposing directions, if the topic is broad or unfamiliar.

3. Iterate: If the user adjusts or rejects directions, refine the proposal until the user explicitly confirms. Do not proceed to Step 3 without confirmation.

---

Step 3: Parallel Web Research

Once the division direction is confirmed, dispatch multiple sub-agents in parallel — one per confirmed direction — to perform deep web research.

Each sub-agent should:

• Search for authoritative sources, learning maps, and expert opinions on the direction.
• Identify the core knowledge areas within that direction (these become candidates for sub-topics).
• Summarize findings: key concepts, typical sub-domains, commonly recognized categories.

Synthesize all sub-agent results into a unified research report before proceeding.

---

Step 4: Propose Sub-Topics (Human-in-the-Loop)

Based on the research report, propose a concrete list of sub-topics to create under the target topic.

Present them clearly:

Based on the research, here are the proposed sub-topics for "[Topic]":

Under direction "Core language fundamentals":
  - HTML & Semantics
  - CSS & Layout
  - JavaScript Fundamentals

Under direction "Frontend frameworks":
  - React
  - Vue
  - Angular

Do you want to adjust, remove, or add any sub-topics before I create them?

Iterate with the user until the list is explicitly confirmed. Do not proceed to Step 5 without confirmation.

---

Step 5: Parallel Sub-Topic Creation

Once the sub-topic list is confirmed, dispatch multiple Create Topic agents in parallel — one per sub-topic — using the Create Topic Skill.

Each agent receives:

• Parent directory: the target topic path (from Step 1)
• Sub-topic name
• One-sentence definition (derived from research)

Skip the "Find Best Placement Directory" step inside Create Topic — the parent is already known.

---

Step 6: Output Result

After all sub-topics are created, output the resulting directory tree:

✅ Dive-into complete for topic: [Topic]

[topic_path]/
├── README.md
├── FAQ.md
├── [subtopic-1]/
│   ├── README.md
│   └── FAQ.md
├── [subtopic-2]/
│   ├── README.md
│   └── FAQ.md
└── [subtopic-3]/
    ├── README.md
    └── FAQ.md

Also update the target topic's README.md Sub-topic Index to include links to all newly created sub-topics (if not already updated by Create Topic agents).

---

Acceptance Criteria

• [ ] Target topic is located via Retrieve Topic subskill
• [ ] Research direction is negotiated and confirmed by user before any research begins
• [ ] Web research is dispatched in parallel across confirmed directions
• [ ] Sub-topic proposals are derived from research and confirmed by user
• [ ] Sub-topics are created in parallel using Create Topic subskill
• [ ] Target topic's README.md Sub-topic Index is updated
• [ ] Final directory tree is outputted

Helper Tools

• Web search: use the web_search tool to research each direction
• Directory tree: tree -L 3 <topic_path>
• Parallel dispatch: invoke multiple sub-agents simultaneously for Steps 3 and 5