Effector. Type-check AI agent skills before runtime.

AI agents are getting more capable, but the skill layer is still brittle. Effector helps builders scaffold and validate SKILL.md early so capabilities become safer, more inspectable, and easier to compose. We’re starting with OpenClaw workflows and a practical wedge instead of leading with a giant platform story.

View on GitHub Read full README

Starting with: OpenClaw SKILL.md Validation

You no longer have to

Discover broken skills only after runtime.
Author agent capabilities as ad-hoc files with no structure.
Guess whether a skill is reviewable or reusable.
Treat capability building as a pile of one-off glue code.

How it works

Scaffold skills from a clearer starting point.
Start with structured SKILL.md workflows instead of blank files and TODOs.
Validate before runtime.
Catch mismatches and broken assumptions earlier, before capabilities fail inside a live agent workflow.
Make capabilities more inspectable.
Build toward skills that are safer to review, easier to reason about, and easier to compose.

Problem

1. Agent runtimes are improving.
2. Skill authoring is still brittle.
3. Errors often appear too late.

Wedge

1. Scaffold SKILL.md.
2. Validate skills early.
3. Inspect capabilities before runtime.

Direction

1. Start with OpenClaw workflows.
2. Improve reviewability and reuse.
3. Grow toward a broader capability layer.