> For the complete documentation index, see [llms.txt](https://surecloud.gitbook.io/surecloud-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://surecloud.gitbook.io/surecloud-docs/documentation/gracie/personas-skills-and-agents.md).

# Personas, Skills and Agents

Understand how Gracie personas, skills, and agents differ, and when to use each one.

## Personas, Skills and Agents

Gracie works through three building blocks.

Each one solves a different job.

* **Personas** give expert judgement.
* **Skills** do a defined job in chat.
* **Agents** run that job inside a workflow.

### Choose the right building block

#### Personas

Use a persona when you want a specialist view before you decide.

Personas answer from the perspective of a senior GRC practitioner.

They stay read-only.

They ground their answer in your real records.

They take a position instead of listing abstract possibilities.

For the current persona catalogue, see [Skill and Persona Reference](/surecloud-docs/documentation/gracie/personas-skills-and-agents/skill-and-persona-reference.md).

#### Skills

Use a skill when you want Gracie to do a specific job now.

Skills can summarise, review, draft, classify, map, or create work for you.

Gracie can select a matching skill from plain language.

You can also invoke a skill directly when it is exposed in chat.

Any skill that creates or updates records shows a draft first.

Nothing is written until you confirm.

See [Skills](/surecloud-docs/documentation/gracie/personas-skills-and-agents/skills.md) for how skills are managed and [Write and Test Custom Skills](/surecloud-docs/documentation/gracie/personas-skills-and-agents/write-and-test-custom-skills.md) for authoring guidance.

#### Agents

Use an agent when the work should run inside a workflow.

An agent is attached to a stage, transition, or button.

It runs when the process reaches that point.

Agents use the same skills library as Gracie in chat.

They also run with an assigned permission context.

See [Agents](/surecloud-docs/documentation/gracie/personas-skills-and-agents/agents.md) for setup and deployment.

{% hint style="info" %}
The simplest distinction is **invoked** versus **deployed**. Skills are invoked in the moment. Agents are deployed into a process.
{% endhint %}

### How they work together

A common flow looks like this:

{% stepper %}
{% step %}

#### Consult a persona

Ask for a specialist view on a risk, project, control, or incident.
{% endstep %}

{% step %}

#### Run a skill

Use Gracie to draft, review, or analyse the work in front of you.
{% endstep %}

{% step %}

#### Deploy an agent

Automate the repeatable part so it runs inside the workflow next time.
{% endstep %}
{% endstepper %}

### Go deeper

<table data-view="cards"><thead><tr><th>Topic</th><th>Summary</th><th data-card-target data-type="content-ref">Target</th></tr></thead><tbody><tr><td><strong>Skills</strong></td><td>Understand how skills trigger, where they run, and how admins manage them.</td><td><a href="/surecloud-docs/documentation/gracie/personas-skills-and-agents/skills.md">Skills</a></td></tr><tr><td><strong>Write and Test Custom Skills</strong></td><td>Build tenant skills that trigger reliably, stay grounded, and confirm before writing.</td><td><a href="/surecloud-docs/documentation/gracie/personas-skills-and-agents/write-and-test-custom-skills.md">Write and Test Custom Skills</a></td></tr><tr><td><strong>Skill and Persona Reference</strong></td><td>Browse the current Gracie persona, skill, and knowledge base catalogue.</td><td><a href="/surecloud-docs/documentation/gracie/personas-skills-and-agents/skill-and-persona-reference.md">Skill and Persona Reference</a></td></tr><tr><td><strong>Agents</strong></td><td>Define reusable AI agents and attach them to workflow stages, transitions, and buttons.</td><td><a href="/surecloud-docs/documentation/gracie/personas-skills-and-agents/agents.md">Agents</a></td></tr></tbody></table>