Routed: Open-Source Local Skill Router for AI Agents Runs in Under 20ms on CPU
iAmQubick · reddit · 2026-09-06
Routed (MIT, open source) is a local-first hybrid router for agent skills that runs fully offline on CPU: lookups complete in under 20ms with zero tokens and no network calls.
It replaces stuffing skill definitions into system prompts (context bloat) or paying for an LLM router turn (1000+ tokens, 2s+ latency) with a 4-part scoring pipeline: quantized ONNX dense embeddings (60%, Arctic Embed S / MiniLM), Okapi BM25 (25%), exact/alias matching (10%), and recency/usage metadata (5%). Auto-installs adapters for Cursor, Claude Code, OpenCode, Antigravity and Codex; /route triggers the best skills, including compound intents. Prebuilt binaries for macOS, Linux and Windows.
More from coding & agent
- We sandboxed our AI agent — it escaped one second later — prasenx · 2026-09-06
- Median OpenAI researcher reportedly spends $601.25/day on coding agents — scaling01 · 2026-09-06
- Andrew Ng: don't let coding agents run for hours — five skills for agentic coding — DeepLearningAI · 2026-09-06
- Researcher calls on RL teams to add refactoring and deletion tasks to agent training — kuza55 · 2026-09-06
- OpenAI Astra reshapes AI-CAD: a breakthrough weekend of complex 3D modeling — burhop · 2026-09-06
- OpenAI Agents Coordinated via a Wiki Page While Ignoring the Human Moderator — PeterHndrsn · 2026-09-06