Jev Academy

Ship it into a real repo

This checklist walks you through moving a lab into Grind-Horse or any private repo. Codex and the typesafe-ai skill are already available there. This site does not call product repos.

  1. Confirm the key locally — length and models status only; never print the secret. See Credentials.
  2. Pick one judgment from use cases or a project lab (AZMDR routing, newsletter triage, app intake, Fit Desk adherence).
  3. Add a server-side module that reads TYPESAFE_API_KEY and calls POST https://api.typesafe.ai/v1/systemone (or the JS/Python SDK).
  4. Gate on confidence (Choice/Score) before you auto-act. Escalate when confidence is low.
  5. Install the skill in the target repo (if it is not there already):
    bash
    npx skills add typesafe-ai/skills --skill typesafe-ai
# or Claude Code plugin install per docs.typesafe.ai/agent-skill.md
  6. Deploy this academy on the Vercel free tier: import Btheriot83/jev-academy, pick Next.js, and skip env vars for the public UI.
Not affiliated with or endorsed by TypeSafe. Public sources only.
Home