LH Knowledge Base gives each site a structured, AI-maintainable knowledge base: stable, materialised knowledge (concepts, workflows, conventions, platform notes) stored as hierarchical entries that people and AI agents can both read and write.
Every entry is also served as plain Markdown at its own .md URL, and folders get a generated index.md, so an agent can navigate the whole knowledge base the same way it would read a documentation site. Full create, read, update and list access is available through MCP abilities.
Key features
- Hierarchical entries, typed as concept, workflow, ability or platform.
- Markdown versions of every entry and folder index at predictable URLs.
- MCP abilities to create, read, update, list and trash entries.
- Visibility is filterable, so it can be tied to membership or capabilities.
