Awesome Search Kg Hubs
Documentsawesome-search-kg-hubsAwesome Search Kg Hubs is an Agent Skill for documents published in the frutik/awesome-search repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.
Audit hub notes in the Awesome Search Obsidian vault — rank notes by inbound link count and compare against existing Topics/ notes. Surfaces mismatches where high-traffic concepts lack a proper topic page. Use for periodic graph structure reviews.
How to use this skill
Bring this guide into your coding agent with a prompt tailored to the tool you use.
- Open your project in Codex.
- Copy and paste the prompt below.
- Review files and risks before approving any action.
Inspect the Agent Skill at https://github.com/frutik/awesome-search/blob/HEAD/claude-skills/awesome-search-kg-hubs/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/awesome-search-kg-hubs, but wait for my approval before writing files or running scripts.
Copying this prompt does not install or run the skill. Review third-party files before use.
Awesome Search KG — Hub Audit
Use this skill to identify the true hub notes in the vault by inbound link count, then compare them against the existing Awesome Search/Topics/ notes. A mismatch — where a concept note has more inbound links than its topic page — is a restructuring signal.
When to use
- After a significant batch of new articles have been processed
- When the Topics/ folder feels out of sync with the actual content
- Periodically (e.g. monthly) as part of graph health maintenance
Vault Access
Always use the Obsidian MCP server for all vault operations — never access the vault directly via the filesystem (no Read/Write/Bash file tools). The MCP server is the only sanctioned way to read, search, create, and edit notes.
Workflow
1. Count inbound links for all notes
For every note in Awesome Search/, count how many other notes contain a [[wikilink]] pointing to it.
Focus on:
Awesome Search/Concepts/— these are the most likely hubsAwesome Search/Topics/— these should be hubs but sometimes aren'tAwesome Search/People/— high-link people may need richer notes
2. Produce the top-N ranking
Output the top 20 notes by inbound link count:
HUB RANKING — Awesome Search KG
================================
Rank | Note | Folder | Inbound Links
-----|-----------------------------|-----------|--------------
1 | Hybrid Search | Concepts | 34
2 | BM25 | Concepts | 28
3 | Search Evaluation | Concepts | 25
...
3. Compare against Topics/
For each top-ranked concept, check:
- Does a corresponding
Topics/note exist? - If yes: does the topic note have fewer inbound links than the concept? (mismatch)
- If no: should one be created?
Flag mismatches:
MISMATCH REPORT
===============
- "Hybrid Search" (Concepts, 34 links) > "E-commerce Search" (Topics, 12 links)
→ Consider elevating Hybrid Search to a Topic note
- "LLM as Judge" (Concepts, 18 links) — no Topic note exists
→ Candidate for new Topic: "LLM-Based Evaluation"
4. Flag under-linked Topics
Also check the reverse: existing Topics/ notes with very few inbound links (< 3). These may be:
- Correctly low-traffic niche topics (OK)
- Topics that haven't been linked from articles yet (needs linking)
- Topics that are redundant with a concept note (consider merging)
5. Recommendations
For each mismatch or gap, produce one of:
- Elevate: promote a concept note to a topic note (or create a new topic note)
- Link: existing topic note just needs more article links added to it
- Merge: topic and concept overlap significantly — consolidate
- Monitor: not actionable yet but worth watching
Do not make changes automatically — present recommendations for user approval.