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

# Approval & Access

Control how much authority Agents have when they take actions on your behalf.

The **Approval & Access** setting determines whether an agent pauses for confirmation before governed actions or can proceed automatically when workspace policy permits. Configure this before starting an agent from the control beneath the prompt.

See also [Governance](https://docs.countercyclical.io/agents/governance) and [Run Inspection](https://docs.countercyclical.io/agents/orchestrations/inspection).

### Ask for Approval

**Ask for approval** is the default, human-in-the-loop mode.

The agent can research, analyze, and prepare work normally, but pauses before governed actions that require your authorization. You can review the proposed action and decide whether it should proceed.

Use this mode when you want to retain direct control over consequential actions, are testing a new agent or workflow, or want to review changes before they are committed.

### Approve for Me

**Approve for me** allows the agent to automatically approve eligible actions on your behalf.

Automatic approval remains subject to **workspace policy**. Selecting this option does not give the agent unrestricted authority or allow it to bypass permissions and governance controls.

This mode is useful for trusted, repeatable workflows where requiring manual confirmation for every eligible action would add unnecessary friction.

### Full Access

**Full Access** provides the highest available level of agent authority.

This option is available only to **workspace owners and admins** when enabled by workspace policy. If your role or workspace policy does not permit Full Access, the option appears unavailable.

Full Access should be reserved for agents and workflows where broad autonomous execution is intentional and appropriate.

### How Approval Interacts with Permissions

Approval and access settings do not replace the workspace permission model. An agent's effective authority is constrained by the combination of the selected approval mode, your own access, applicable workspace policies, and the permissions governing the resources or actions involved.

As a result, **Approve for me** should be understood as permission to automatically approve *eligible* actions—not as permission to override workspace controls. Likewise, **Ask for approval** introduces a confirmation checkpoint but does not make an otherwise unauthorized action available.

{% hint style="warning" %}
**Choose a Review Mode**

For governed environments, **Ask for approval** provides the strongest human review checkpoint, while **Approve for me** and **Full Access** progressively reduce manual intervention where workspace policy allows it.
{% endhint %}

#### Related Pages

{% content-ref url="/pages/Rl2WZFJmGl0gjAjkDEyA" %}
[Governance](https://docs.countercyclical.io/agents/governance.md)
{% endcontent-ref %}

{% content-ref url="/pages/x5NcwZqAn0tjF3KR1AXS" %}
[Inspection](https://docs.countercyclical.io/agents/orchestrations/inspection.md)
{% endcontent-ref %}

{% content-ref url="/pages/SEBcQQoACGG4eLh49Wda" %}
[Agents Overview](https://docs.countercyclical.io/agents/agents-overview.md)
{% endcontent-ref %}
