Skill library
The instructions your agents run on, in one place.
Skills are the reusable procedures your team gives its agents. Keeping them in one approved library, rather than in one person’s local folder, means improving a skill improves every agent that reads from it.
Approved and versioned
A skill moves from draft through review to approved, and every version is kept. Agents connect to the library and read the current published version, so an improvement lands everywhere at once and a bad change can be traced to the version that introduced it.
- Draft, review and approval before a skill goes live
- Versions published deliberately, with the history kept
- Agents read the library directly rather than a copied file
Health checks catch a skill going stale
A skill that references a renamed file or a retired process fails quietly, one agent run at a time. Health checks surface that before it becomes the reason a deliverable was wrong, and stewardship signals show which parts of the library nobody is maintaining.
Analytics on what actually gets used
Usage events show which skills run, how often, and what they cost against the human time on the same workstream, so the ones nobody invokes are visible.
- Usage counts per skill
- Skill economics against human time on the same work
- Stewardship signals for the parts going unmaintained
Start counting all the work.
30-day free trial. Bring the time history you already have.