How we measured a markdown knowledge graph as agent memory — lost to grep, rebuilt our search and editing primitives, retracted our own best number, and ended with a $4.50 curator whose store reads back at 96% of a hand-built ceiling in a single call.
https://t.co/UIzTA0uMvF
The ultimate "second brain" for AI agents to perfectly understand your context 🧠
Check out "IWE", a next-gen Markdown knowledge management system built in Rust!
It ditches traditional folders, connecting notes via links to build a flexible "Knowledge Graph".
With native MCP support, tools like Claude and Cursor can directly explore, read, and write to your personal knowledge base ⚡
https://t.co/LmlKLfhvSH
#LLM #AI #MCP #Markdown #KnowledgeGraph #Rust #ContextEngineering
iwe processes 20,000 documents in under a second.
No database, no indexing step — just plain markdown files parsed into a graph and ready for instant search and filtering.
Benchmarks: https://t.co/B7qyFOwnOg
iwe schema — scan your knowledge base and get the full frontmatter structure: field names, types, coverage, distinct values, and distributions.
One command to understand how your documents are shaped.
https://t.co/oqE3kGMcnq
iwe v0.1.0 — query your markdown knowledge graph
YAML filter expressions, graph traversal (--includes, --references), frontmatter projection into JSON, new count & update commands.
Your docs are a queryable database now.
https://t.co/lydbxm0RVt
Your markdown notes are a graph. Now you can query it like one.
iwe just got structural set selectors — slice your knowledge base by intersection of parent docs:
iwe find --in projects/alpha --in topics/security
→ everything security-related inside the Alpha project.
New in iwe: configurable markdown formatting.
Your formatter, your style. Choose emphasis chars, list markers, fence chars — iwe normalizes every note in your vault to match.
https://t.co/oqE3kGMcnq
#markdown#pkm#rust
iwe now has an MCP server — give your AI agents direct access to your knowledge graph.
13 tools for search, retrieval, creation, and refactoring. Built-in prompts. Live file watching.
https://t.co/oqE3kGMcnq
#MCP#AI#rust