Secure secret handling for Claude Code: request, store, use, and reveal secrets without them ever entering the chat.

Synapse is a high-performance llama.cpp fork built on ik-llama.cpp and rk-llama.cpp, focused on efficient AI inference and deployment on edge devices.

A visual comparitive demonstration of Engram vs traditional stateless inference.

HTTP-client benchmark suite for Engram stateful inference. Runs against a server URL; results to s3://engram/benchmarks/.

Harness the power of Engram on your existing Langchain - Langraph applications with our custom provider

Stateful Mamba inference provider for LiteLLM. Snapshot-based persistent state with zero application code changes.

Automated Wiki Creation and Agentic Retrieval System based on Andrej Karpathy's LLM-Wiki Spec and OpenAI's Symphony

A knowledge graph built from plain markdown files — no database, no lock-in. YAML frontmatter carries entities, relationships, and temporal metadata so every file is simultaneously a readable docum…

Use Engram with your existing Vercel AI SDK Projects with this Vercel AI Provider Package

Use Engram's stateful inference engine in your existing Pydantic AI applications without refactoring!

Intel's Edge Device Templates, Libraries and Mini Apps refactored for use with Rockchip based Edge Devices.