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

# Corunner Skills: How Agents Reason and Learn

> Skills are the reusable thinking methods that shape how Corunner agents reason. Builtins are shared; admins can layer org-specific guidance on top.

Skills are how Corunner's agents reason: reusable thinking methods that describe an approach, a domain, or a framework. Builtin skills are shared across all customers and change through pull-request review. Admins can layer **org-specific guidance** on top of any builtin skill without editing the shared version.

## Kinds of skills

The Skills page in the web app filters by kind:

* **Domain** — how to reason about a specific business area.
* **Task** — how to complete a recurring task.
* **Framework** — a named method or framework the agent should follow.
* **Provider** — how to reason about a specific integration or provider.
* **Agent** — guidance scoped to a single agent in the roster.
* **All** — the full list.

## Layer an org-specific override

<Steps>
  <Step title="Open Skills">
    In the Corunner web app, go to **Skills**.
  </Step>

  <Step title="Filter to the kind you want">
    Choose a kind (Domain, Task, Framework, Provider, Agent) or leave it on **All**.
  </Step>

  <Step title="Open a skill">
    Click into the skill to see its current guidance. Your existing override is pre-filled if one exists.
  </Step>

  <Step title="Save your override">
    Add your organization-specific guidance and save. Corunner runs a safety scan before accepting the override. Empty or flagged content is rejected.
  </Step>
</Steps>

## Revert an override

Reverting an override is destructive: it drops your tenant customization and returns the skill to the shared builtin behavior. Corunner asks for confirmation before reverting.

<Note>
  Authoring and reverting skill overrides is admin-only.
</Note>

## Related

<CardGroup cols={2}>
  <Card title="Agent Roster" icon="robot" href="/agents/roster">
    See which agents skills apply to.
  </Card>

  <Card title="Governance" icon="scale-balanced" href="/admin/governance">
    Policies that decide what agents can do on their own.
  </Card>
</CardGroup>