Ax0n

Every time you start a new AI chat, you re-explain everything. Ax0n fixes that.

Like the axons in your brain that carry signals between neurons, Ax0n carries context between your sessions. It runs silently in the background, captures what you work on, and automatically surfaces relevant context the next time you prompt your AI assistant with a related topic. No copy-pasting, no manual notes, no setup per project.

Why Ax0n: Most AI assistants start from zero every chat. You re-explain your stack, repeat past decisions and tradeoffs, and recap work you did last week. Ax0n remembers what you've already established and surfaces it when it's relevant — so you stop paying for the same context twice. Saving time AND tokens.

How it works

Manual capture: Select any text from a file and press Cmd+Shift+C to save it as a memory. All memories appear in the Ax0n sidebar in your activity bar.

Local embeddings: Each capture is embedded using all-MiniLM-L6-v2 (via @xenova/transformers) and stored in a local SQLite database. Nothing leaves your machine.

MCP server: Ax0n exposes a search_memories tool and a save_memory tool over the Model Context Protocol. Any MCP-compatible client can call them.

Automatic context injection: The repo ships with rules that tell the AI what to do automatically. No prompt engineering required on your end.

At the start of every conversation, the AI calls search_memories based on what you ask and anything relevant comes back as context before it responds. After every substantive response (a fix, implementation, diagnosis), it calls save_memory with a 2–3 sentence summary of what it did and why. Over time, context accumulates across sessions.

Ax0n works in both Cursor (via .cursor/rules/ax0n-memory.mdc) and Codex (via AGENTS.md). Both files ship with the repo and are already active when you clone it.

Setup

1. Install the extension

In VS Code or Cursor, open the Extensions view and search for Ax0n (publisher patel-haley), then install.

Alternatively, download a .vsix from Releases and run Extensions: Install from VSIX in the Command Palette.

2. Register the MCP server (pre-filled snippets)

Open the Command Palette (Cmd+Shift+P / Ctrl+Shift+P) and run the command for the client you use. Each command copies a ready-to-paste config that already points at mcp-server.js inside your installed Ax0n extension (and picks a sensible node command). Merge into existing MCP config if you already have other servers.

After pasting, restart Cursor or Codex. For Codex, run codex mcp list and use /mcp in the chat to confirm ax0n is listed.

If the MCP server won’t start: The copied config often uses plain node as the command. Cursor and Codex launch that process with a minimal environment (not the same as your terminal) so they sometimes can’t find node on your PATH. That’s especially common if you use nvm, fnm, or asdf, because those tools usually add Node only when a shell starts up.

Fix: In a normal terminal, run which node and copy the full path it prints (for example /Users/you/.nvm/versions/node/v22.0.0/bin/node). Edit your pasted config and replace the command value with that full path instead of node. Save, restart Cursor or Codex, and try again.

3. Optional — Ollama summarization

By default, Ax0n embeds your raw captured text. With Ollama, it first summarizes the text into a concise description before embedding, which means the semantic search finds more relevant memories and filters out noise. Recommended if you're capturing long functions or dense code blocks where raw text would produce noisy embeddings.

Install Ollama and pull the default model:

brew install ollama
ollama pull llama3
ollama serve

To use a different model, set ax0n.ollamaModel in VS Code settings (default: llama3). Set it to an empty string to disable summarization even if Ollama is running.

What it captures

• Manual selections via Cmd+Shift+C or Ax0n: Capture Selection
• AI chat summaries written by the MCP save_memory tool after substantive responses

Managing memories

The Ax0n sidebar (activity bar icon) shows all saved memories with their source and timestamp. Memories saved via MCP from AI chats appear automatically.

Available commands (open Command Palette with Cmd+Shift+P):

• Ax0n: Search Memories — semantic search across all memories
• Ax0n: List Memories — view all memories in the output panel
• Ax0n: Delete Memory — select and delete a specific memory
• Ax0n: Clear All Memories — delete all memories (requires confirmation)
• Ax0n: Show Top Result for Current File — quick relevance check for the active file

Memories are automatically pruned after 30 days if they've never been accessed.

Built with

| | | |---|---| | Embeddings | @xenova/transformers — all-MiniLM-L6-v2, runs fully in-process | | Storage | better-sqlite3 — local WAL-mode database | | MCP | @modelcontextprotocol/sdk | | Language | TypeScript | Fully local. No API keys. No telemetry.