Skip to main content
Agent Skills are portable instruction sets that teach AI coding assistants how to perform Tusk tasks — like analyzing API test deviations, querying recorded traffic, or debugging regressions without leaving your editor. Skills work with 18+ AI agents including Cursor, Claude Code, GitHub Copilot, Cline, and many others.

​
Installing skills

Install all Tusk skills: 
npx skills add Use-Tusk/tusk-skills
Install a specific skill: 
npx skills add Use-Tusk/tusk-skills --skill tusk-drift-analyze
Install to specific agents: 
npx skills add Use-Tusk/tusk-skills -a claude-code -a cursor

​
Available skills

​
Tusk Drift

These skills require Tusk Drift to be set up in your project. See the Drift overview to get started, and Tusk Drift Cloud to connect your service.

tusk-drift-analyze

Run the Drift CI/CD check locally — replays API tests, classifies deviations as intended, unintended, or unrelated, and optionally fixes regressions.

tusk-drift-query

Search, analyze, and debug recorded API traffic using the Tusk Drift CLI query commands — covers spans, traces, aggregations, and schema exploration.

​
Tusk Unit Tests

This skill requires Tusk unit test generation to be enabled on your repo. See CLI & Agent Workflows for the underlying CLI primitives.

tusk-unit-tests

Discover, reason about, and selectively adopt Tusk-generated unit tests for the current branch. Classifies each scenario as fresh or stale against the local checkout, applies the ones worth keeping, and reports feedback back to Tusk.
Browse the full skill catalog and source on GitHub. New skills are added as Tusk products evolve.

​
How skills work

Each skill is a SKILL.md file containing structured instructions that agents load on demand. When you ask your agent to do something Tusk-related, it reads the skill to understand which CLI commands to run, how to interpret the output, and what workflows to follow.