Integrations

Engram SpectreVS Code

Your local AI pair programmer. Spectre lives directly inside VS Code and uses your local Llama 3 model to analyze, refactor, and explain code—without a single byte leaving your machine.

Usage

Select any code snippet, open the Command Palette (Cmd+Shift+P), and run "Spectre: Ask".

Privacy First

Inference happens locally via Docker. No telemetry, no cloud API keys, and zero data egress.

Installation

01Package the Extension

Navigate to the engram-spectre directory and package the extension into a VSIX file:

bash

vsce package

02Install VSIX

Install the generated package directly into your VS Code instance:

bash

code --install-extension engram-spectre-0.0.1.vsix

Troubleshooting

Spectre Error: Connection Refused

Spectre cannot reach your local backend. Ensure the Engram OS is running and accessible.

./start_os.sh