> ## Documentation Index
> Fetch the complete documentation index at: https://docs.corvic.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Agent Skills

> Learn how to define and use agent skills so your agents can invoke reusable, governed capabilities

## Agent Skills

This tutorial shows how Corvic AI supports **agent skills**—reusable capabilities your agents can call with clear inputs and outputs. Skills help you package domain logic, tools, and workflows so agents stay consistent, testable, and easier to govern.

<iframe className="w-full aspect-video rounded-xl" src="https://www.youtube.com/embed/Z670Yyz3Zcw" title="Agent Skills - Corvic AI" frameBorder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowFullScreen />

## What You'll Learn

* What agent skills are and when to use them
* How skills fit into agent design alongside data grounding and tools
* Patterns for composing skills with agents and data apps

## Key Concepts

### Reusable capabilities

Skills encapsulate behavior your agents can invoke repeatedly—whether that is retrieval over a specific corpus, a structured workflow, or integration with an external system—without duplicating prompts or logic across agents.

### Governance and clarity

Well-defined skills make it easier to review what an agent can do, limit scope to approved operations, and iterate on behavior independently of the rest of the agent configuration.

## Working with agent skills

### Step 1: Identify repeatable work

List tasks your agent performs often (summarization over a table, policy checks, formatted handoffs). Candidates with stable inputs and outputs are strong skill boundaries.

### Step 2: Define the skill contract

Specify what the skill accepts, what it returns, and any side effects or data it is allowed to touch. A clear contract makes skills composable and easier to test.

### Step 3: Attach skills to agents

Configure your agent to invoke the right skills during a run. Combine skills with grounded retrieval and your embedding spaces so answers stay tied to enterprise data.

### Step 4: Validate in realistic scenarios

Exercise skills with representative queries and edge cases. Adjust boundaries when you see drift between intended behavior and production usage.

## Benefits

Agent skills reduce duplication, speed up iteration on agent behavior, and make production agents easier to reason about—especially as you scale the number of agents and data apps in your organization.

## Best Practices

Keep skills focused; prefer several small skills over one large catch-all. Document assumptions and failure modes. Reuse skills across agents where behavior should stay identical.

## Related Documentation

<CardGroup cols={2}>
  <Card title="Building Domain Specific Custom Agents" icon="robot" href="/tutorials/custom-agents">
    Learn how to build agents grounded in multimodal, multi-structural data.
  </Card>

  <Card title="Agents" icon="robot" href="/concepts/agents">
    Core concepts for configuring and using agents.
  </Card>

  <Card title="Data Apps" icon="window" href="/features/data-apps">
    Deploy agents and workflows as applications.
  </Card>

  <Card title="Video tutorials overview" icon="list" href="/tutorials/overview">
    Browse all Corvic AI video tutorials.
  </Card>
</CardGroup>

## Additional Resources

* [Watch on YouTube](https://www.youtube.com/watch?v=Z670Yyz3Zcw\&list=PLg5ygoWwy28IOi8xYqjr_HlSBJorIOEvu\&index=1)
* [Full Playlist](https://www.youtube.com/playlist?list=PLg5ygoWwy28IOi8xYqjr_HlSBJorIOEvu)