
Tessl is a developer tool that evaluates and optimizes AI agent skills before you deploy them. It catches bad outputs, hallucinations, and wasted compute early by scanning your skill code through a GitHub URL or directly from your terminal. No signup is required to get started.
Paste any public repository URL, and Tessl automatically scans the skill. The evaluated skill is added to the open registry, making it discoverable and installable by other developers.
Run npx tessl skill review ./ directly in your terminal. This works with private repos and unfinished skills, keeping all results local and visible only to you.
Evaluated skills become part of a public registry where developers can find, install, and build upon each other's work. This creates a shared library of vetted agent capabilities.
"Bad skills burn tokens, produce wrong output, and send you back to square one."
Tessl addresses the core pain point of AI agent development: unreliable skills that waste time and compute. By catching problems early through automated evaluation, it prevents the frustrating cycle of debugging hallucinations and incorrect outputs. The dual approach of public GitHub scanning and private local reviews gives you flexibility depending on your workflow and privacy needs.
You build or integrate AI agent skills and want to avoid the hidden costs of bad outputs. Tessl is especially useful if you frequently pull skills from GitHub and need confidence they'll work correctly, or if you're developing proprietary skills that require private testing before release. The zero-signup access makes it easy to try without commitment.
Other tools you might consider
Loading comments…
Maker
meowbyte
Visit Website
tessl.io/registry/skills/submit/
Project Info
Product Keywords
Alternatives