jev-agent-skill-router

# Jev Agent Skill Router

<img src="assets/jev-agent-skill-router-banner.png" alt="A request enters a typed routing layer and branches to specialist skills, no skill or human review" width="896">

[](https://github.com/GodsBoy/jev-agent-skill-router/actions/workflows/checks.yml)
[](https://docs.python.org/3/)
[](LICENSE)
[](https://docs.typesafe.ai/)

**68 of 72 synthetic requests routed correctly (94.4%), versus 51 of 72 (70.8%) for a lexical baseline.** The final TypeSafe Jev run selected no wrong skills and made no needless loads, but returned three unnecessary reviews and one client-validation failure. Median end-to-end latency was **1,287 ms**, with p95 **1,406 ms**.

These are **exploratory, reused-data results**, measured on 16 September 2026 with pinned `jev-1.13.0`. The routing questions were revised after inspecting an earlier full run. The original policy scored 64/72 overall and 52/60 on its initially untouched test split. Neither run establishes production accuracy or calibrated probabilities. Typed output guarantees shape, not correctness.

[Final report](results/final-v1.1.md) | [Complete final JSON](results/final-v1.1.json) | [Original full run](results/final.md) | [Evaluation protocol](docs/evaluation.md)

## Why this exists

[Hermes Agent](https://github.com/NousResearch/hermes-agent) is an extensible agent platform whose skills bring task-specific procedures to a wide range of work. The motivating Hermes profile exposes **443 unique skill names**. That scale makes skill selection a useful problem to study separately from skill execution, with uncertainty visible to the caller.

This is an optional routing layer, not a replacement for Hermes. The private catalogue is not included or used in the public benchmark. The benchmark uses **24 synthetic skills