Propose-then-execute · 100% local possible · open-source

You speak.
The agent structures.

An AI-powered second brain that writes your knowledge base for you. Drop a raw thought — CortX identifies entities, creates Markdown files, weaves wikilinks. Git commits everything automatically.

Download for Windows Build from source
macOS & Linux coming soon · free · open-source (ISC)
CortX main UI — chat, knowledge graph, and agent activity in a single resizable window
// THE PROBLEM

Obsidian, Notion, Logseq — they all assume you will structure your own thinking.

The #1 barrier to a real "second brain" isn't the tool. It's the cognitive overhead of maintenance: naming, tagging, linking, filing.

CortX brings that cost to zero. You type raw text. The agent does the rest.

// FEATURES

Everything a second brain should do

An agent that writes your files — not a chatbot that just talks.

Structuring Agent

Type raw text — the agent identifies entities (people, companies, concepts), creates or updates Markdown files, and adds cross-links automatically.

Interactive Knowledge Graph

Real-time Cytoscape · cose-bilkent graph. Nodes colored by type (person, company, domain, project) with filtering, search, and double-click exploration.

Hybrid Search (RAG)

FTS5 full-text + semantic embeddings. The agent retrieves relevant files before every response, with multi-hop expansion across wikilinks.

Document Library

Import PDF, DOCX, XLSX, PPTX. Text extracted via docling, chunked, vectorized. Auto-linked to entities via [[wikilinks]].

Idle Mode — Passive Insights

In the background, the agent explores your knowledge base: hidden connections, contradictions, gaps, recurring patterns. Archive any insight as a note in one click.

Web & Wikipedia

/wiki imports a Wikipedia article, /internet fetches a URL or runs a DuckDuckGo search — all injected into the agent's context window.

Automatic Git Versioning

Every accepted action = a Git commit. Full history, one-click undo (git revert), and a built-in audit log for every agent action.

Propose-then-Execute

Every action is previewable (before/after diff) and requires your explicit approval. Nothing is ever written to disk without your sign-off.

EN / FR Bilingual

Full interface in English and French. Web search and Wikipedia fetch run in the active language with automatic English fallback.

// DEMO

One raw note → 3 files updated

The agent understands context, decomposes entities, proposes changes. You approve.

Your input

"Coffee with Sarah Kim. She left OpenAI to co-found Redline, an AI safety startup. Mentioned that Anthropic's Constitutional AI V2 paper drops Q3 — apparently it addresses the scalable oversight problem directly."

Agent proposal
~ Network/Sarah_Kim.md
✏️ Status: OpenAI → Redline (co-founder)
➕ Interaction: coffee 04/22/2026
+ Companies/Redline.md [NEW]
📄 AI safety startup · contact Sarah Kim
~ Domains/AI_Safety.md
➕ News: Constitutional AI V2 paper Q3
➕ [[Anthropic]] · scalable oversight

From a single raw note, the agent identifies 3 people, 3 companies, 1 domain, modifies 2 files, creates 1 new file, and maintains all cross-links — in under 3 seconds.

// SCREENSHOTS

A closer look

Agent proposal — before/after diff
Diff preview — every file change is visible before you approve it.
Interactive knowledge graph
Knowledge graph — real-time visual exploration of entity relationships.
Library and Idle Mode insights
Library & Idle Mode — PDF/DOCX/XLSX/PPTX imports alongside background-generated insights.
// YOUR DATA, YOUR MODEL

Run local or in the cloud

Plug in a local LLM for full privacy, or an API key for maximum reasoning power.

100% Local

Ollama · llama.cpp · LM Studio. Zero data leaves your machine. No account, no subscription.

  • ✓ Complete privacy — air-gapped capable
  • ✓ No API key required
  • ✓ Free, unlimited usage
  • ✓ Works fully offline
Recommended: Qwen3 14B · Gemma3 12B · Mistral Small 24B

Cloud API

Claude (Anthropic) or OpenAI-compatible. Best reasoning quality for complex knowledge graphs.

  • ✓ Claude Opus 4.7 / Sonnet 4.6
  • ✓ Real-time streaming
  • ✓ Extended reasoning on complex notes
  • ✓ Key stored locally, never uploaded
Compatible: Anthropic · OpenAI · any /v1/chat/completions endpoint
// WHY NOT OBSIDIAN?

How CortX is different

Mem.ai Khoj Obsidian + AI CortX
Agent writes the files✕✕✕✓
Local Markdown files✕✕✓✓
100% local LLM possible✕✓✕✓
Knowledge graph built-in✕✕✓ (plugin)✓
Propose-then-execute✕✕✕✓
Git versioning + undo✕✕✕✓
// INSTALL

Up in 3 commands

Node.js 20+ · Git · a local or cloud LLM

terminal 
# clone
git clone https://github.com/gcorman/CortX.git && cd CortX

# install
npm install

# run
npm run dev       # dev mode with HMR
npm run dist      # build Windows installer (NSIS)
STEP 1
Set base folder

Pick where Markdown files live (default: ~/Documents/CortX-Base/).

STEP 2
Pick a LLM

Ollama / llama.cpp (local, no key) or Anthropic / OpenAI (API key).

STEP 3
Start typing

Drop raw text. The agent proposes. You approve. Git commits.

Note: If better-sqlite3 fails to load after install, run npm run rebuild to recompile native bindings against your Electron version.

Ready to stop filing?

You speak. The agent structures. Git versions. Everything stays on your machine.

View on GitHub Report a bug