LLM Wiki Agent
An agent skill that ingests documents to build a persistent interlinked wiki.
Drop source files into your project and let AI build and maintain an interlinked wiki.
llm-wiki-agent runs inside coding assistants like Claude Code, Codex, or Gemini CLI to ingest documents and maintain an ongoing markdown wiki. It is built for researchers, executives, and professionals who want to compile deep reference material without manual note-taking.
What it is
The agent reads raw markdown, PDFs, office files, and transcripts, extracts core entities and concepts, updates a living overview, and flags contradictions automatically. It compiles a persistent wiki structure complete with a local vis.js knowledge graph and missing-source lint reports.
Who it's for
Suits knowledge workers, researchers, and technical writers managing heavy reading workloads across multiple sources. Does not suit users looking for a traditional point-and-click GUI app without a terminal or coding assistant workflow.
Key features
Automatic Format Conversion
Converts PDF, DOCX, PPTX, and other formats to markdown on the fly during ingest.
Contradiction Flagging
Flags conflicts between newly ingested sources and existing wiki claims immediately.
Interactive Knowledge Graph
Generates a self-contained graph.html file mapping explicit wikilinks and inferred connections.
Automated Linting Reports
Scans your wiki for orphan pages, broken links, and data gaps with suggested reading.
Up and running in 3 steps
Our honest verdict
What's great
- Maintains an ongoing structured wiki instead of forcing you to search raw chat logs or static folders.
- Auto-creates dedicated entity and concept pages every time new sources mention them.
- Combines deterministic wikilinks with semantic AI-inferred relationships for the knowledge graph.
- Works directly inside existing coding assistants like Claude Code, Codex, and Gemini CLI.
Where it falls short
- Requires you to run a compatible coding agent tool such as Claude Code, Codex, or Gemini CLI.
- Managing a growing wiki file structure requires discipline with the agent prompts and schema configs.
- Relies on markitdown for conversions, meaning complex document layouts may require manual cleanup.
Pricing & license
Free under its open source license, with no paid hosted option provided by the repository. The only cost is running it yourself via your preferred AI agent.
Open remote AI-training roles
870 open remote AI-training roles, updated November 14, 2026; most pay $20–$170 an hour. Zekai lists them free; the application goes to the employer’s platform.
Alternatives
These reviewed AI tools do the same job instantly — from our scored directory.
Use it, or have us set it up
FAQ
Do I need an API key to run this agent?
No separate API key is needed beyond the credentials used by your underlying coding agent tool.
Can I view the knowledge graph in a web browser?
Yes, building the graph generates a self-contained graph.html file that opens in any browser.
What file formats can the agent ingest?
It supports markdown, PDF, DOCX, PPTX, XLSX, HTML, TXT, CSV, JSON, XML, RST, EPUB, and more.
