Skip to content
Get packs

Understand Knowledge Wiki Graph

Turns a Karpathy-style markdown wiki into an interactive knowledge graph dashboard.

Source Download skill
Repository
Egonex-AI/Understand-Anything
Stars
89,744
Licence
MIT
Files
3 files

What it does

When installed, the agent can take a folder of linked markdown notes (a Karpathy-style LLM wiki with an index.md, wikilinks like [[topic]], and optionally raw sources and a schema file) and map how the notes connect. It runs a Python script to parse the files and resolve links, may dispatch sub-agents to infer implicit relationships and topic clusters, then merges everything into a knowledge-graph JSON and an interactive dashboard. Think of it as turning a personal or team wiki into a visual map of its ideas.

Great for

  • People who keep a personal or team markdown wiki built with an LLM and want a visual map of it
  • Researchers or students organizing many linked notes by topic
  • Developers maintaining an internal documentation wiki who want to spot orphaned or unresolved links

Skip it if

  • Folders that are not Karpathy-pattern wikis: the detection script exits with an error for these, so the skill adds nothing.
  • Wikis with sensitive content sent to an LLM: article content is passed to sub-agents for analysis, so confidential notes leave the local folder through the model provider.
  • Raw PDFs or binary sources: source nodes are only filename and size, so their contents are not analysed.

Try asking

  • “Map the knowledge graph of my wiki folder at ~/notes/wiki”
  • “Analyze this Karpathy-pattern wiki and show me the implicit relationships between articles”
  • “Generate an interactive dashboard of how the topics in this knowledge base cluster”
Step-by-step guide: how to use Understand Knowledge Wiki GraphInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 3 original files, the MIT licence and an attribution note.

    understand-knowledge.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip understand-knowledge.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip understand-knowledge.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Point the skill at the wiki root directory containing index.md so categories and layers are built correctly.
  • Review the detected-article count and unresolved wikilink count before trusting the graph; unresolved links usually mean naming mismatches.
  • Back up the wiki before running: the skill writes .ua/ (or .understand-anything/) data folders inside the target directory and deletes its intermediate folder when done.

What's inside

  • merge-knowledge-graph.py
  • parse-knowledge-base.py
  • SKILL.md

Redistributed unmodified from Egonex-AI/Understand-Anything under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.