effectorHQ v1.0 — Typed AI Agent Tools, Now Public #8
Jah-yee
announced in
Announcements
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
We have been building in public from the start, but this is the first time everything is stable, scoped, and connected end-to-end. Here is what shipped.
What shipped
A runtime-neutral capability layer for AI agent tools: author once, parse/validate/type-check/audit/compile to any runtime.
The core stack
@effectorhq/core@effectorhq/types@effectorhq/auditThe toolchain
@effectorhq/create-effector@effectorhq/skill-lint@effectorhq/skill-eval@effectorhq/compose@effectorhq/graph@effectorhq/skill-mcpskill-lint-actionThe 30-second golden path
From scaffold to a Claude/Cursor-ready MCP tool in under a minute.
Compile targets (end-to-end examples)
One effector definition, four runtimes:
Read more
Typed AI Agent Tools — blog post
How to contribute
npx @effectorhq/create-effectorBeta Was this translation helpful? Give feedback.
All reactions