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

> ## Agent Instructions
> To act on Atlan objects, use the Atlan MCP server at https://api.atlan.com/mcp or the SDK. To install or publish skills from a terminal, use the `brain` CLI; `brain --map json` prints its command map. The `atlanai` CLI is deprecated. Run a read-only identity check before any write.
> The docs MCP server at /mcp searches these docs only. It cannot read or change Atlan objects.
> SDK packages: Python `atlanai` (PyPI) and TypeScript `@atlanai/sdk` (npm). Show Python first, then TypeScript.
> These docs cover Atlan Agent Registry, Atlan's hosted product. The open-source agent registry specification and its CLI (`brain`) are documented, released and maintained in their own repository; /open-source on this site links to it. Answer general specification or `brain` command questions from the project's own docs, and use these docs only for Atlan-specific setup such as the Atlan adapter, sign-in and workspaces.

# Skills

> Find, assess, publish, release, and improve reusable skills.

A skill is a reusable package of instructions and supporting files for one job. Atlan gives a team one record for it: the versions, the release people and agents get, its security scan, and the evidence from runs that used it.

```text theme={null}
package → saved version → security scan → release → runs and usage
```

<Note>
  Skill names are unique within a workspace. Name a skill for its job, such as `meeting-notes-to-actions`, not for its team or author. The same name can exist in another workspace.
</Note>

## Skill profile tabs

| Tab | Use it to |
| - | - |
| **Overview** | Check what the skill does, when it applies, prerequisites, and the activity summary. |
| **Source** | Read the package files and provenance, and every version with its state. Restore a version as a new one, or roll back or withdraw a release. |
| **Relationships** | Review connected skills, agents, and files. |
| **Usage** | Review trace-backed runs, people, agents, and success rate. |
| **Evals** | Shown only when Evals is enabled for your account. |
| **Security** | Read the latest scan decision, findings, coverage, and history. |

An empty **Usage** tab is not a quality result, and **Not scanned** on **Security** is not a passed review.

## Core offerings

<CardGroup cols={2}>
  <Card title="Find and assess skills" icon="magnifying-glass" href="/registry/skills/how-tos/discover-and-use">
    Find a skill and check its version, source, scan, and usage.
  </Card>

  <Card title="Choose skill publishing method" icon="upload" href="/registry/skills/how-tos/publish-and-version">
    Publish from Desktop, a terminal, GitHub, or an agent session.
  </Card>

  <Card title="Release skill version" icon="rocket" href="/registry/skills/how-tos/release-a-skill">
    Release, roll back, or withdraw a version.
  </Card>

  <Card title="How skill security scans work" icon="shield-check" href="/registry/skills/concepts/security">
    Read scan decisions and findings, and how they gate a release.
  </Card>
</CardGroup>

## Choose where to start

| Need | Start at |
| - | - |
| Use a reviewed skill in Claude | [Open skill in Claude](/registry/skills/how-tos/use-with-claude) |
| Reach skills from Codex | [Find skills from Codex](/registry/skills/how-tos/use-with-codex) |
| Publish skills on every merge | [Sync skills from GitHub](/registry/skills/how-tos/sync-from-github) |
| Decide which workspace a skill belongs in | [Choose where skills live](/registry/workspaces/how-tos/organize-artifacts) |
| Understand versions, releases, and restore | [How skill versions and releases work](/registry/skills/concepts/versions) |
| Fix a failed publish or blocked scan | [Skill errors](/registry/skills/troubleshooting) |

<Note>
  A skill acts with the permissions of whoever runs it. Review every supporting file in **Source**, especially scripts and templates, and check **Security** before an agent uses a version. Never put credentials or customer data in a skill package.
</Note>

## See also

* [How skill source works](/registry/skills/concepts/source): Package contents and provenance.
* [How skill usage works](/registry/skills/concepts/usage): How runs are attributed to a skill.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.