Top Repos· Jev-Based· Harness· Last updated

typesafe-ai / skills

“Agent skills for building with TypeSafe's System One API”

Jev-Based rank #4In Harness #3

TL;DR

typesafe-ai/skills is vendor GitHub — treat it as a TypeSafe source, not a jev.pro artifact. Canonical request facts stay on docs.typesafe.ai.

The author describes it as: “Agent skills for building with TypeSafe’s System One API” That is their claim, not a bench this site ran.

Who it’s for

What’s inside

Top-level entries on GitHub at generation time: .claude-plugin, LICENSE, README.md, skills. Read those before you trust a one-line install.

How to use

  1. Read SKILL.md in the repository. The skill teaches the API; it is not a key booth.
git clone https://github.com/typesafe-ai/skills.git
  1. Install the skill in the repo that will call Jev.
  2. Keep TYPESAFE_API_KEY (or a documented gateway secret) in that environment.

Hosted Jev still looks like a function call — wrapper or not:

{
  model: "jev-1.13.0",
  state: { source: "typesafe-ai/skills", brief: "..." },
  questions: {
    gate: { type: "noul", instructions: "Should we act on this state?" },
    route: { type: "choice", instructions: "Pick one next system.", criteria: ["code", "review", "human"] },
  },
}

Try this

First pinned call

Call typesafe-ai/skills with model jev-1.13.0, a two-sentence support ticket as state, and questions: refund_asked (noul), department (choice: billing|tech|other). Print choice, noul, confidence, and model.

Schema-safe ≠ correct

On a case typesafe-ai/skills gets wrong, write down the schema-legal answer it returned and the human label. Do not “fix” it by asking Jev to generate an explanation.

FAQ

Is this the Jev model?

No repository is the model. Jev is the hosted System One instance. This repo is vendor code around the API.

Can I get a key here?

No. Keys come from TypeSafe’s console or a documented gateway.

Sources

Public repository metadata and TypeSafe documentation. No private claims.