Settings

The Settings tab controls organization-wide feature availability, app access, and reusable content imports. These settings affect the whole organization and should be changed by an owner or administrator.

Feature Availability

The Feature availability section contains:

• Agents: agents are always available; select Default LLM connection and the separate organization Default agent.
• Public Chat: controls whether public chat surfaces can be enabled.
• Webhooks: controls availability of webhook entry points.
• API: controls API access and API-key management, including personal keys.
• Recordings: controls the Recordings module.
• Quick actions: enables quick actions and selects their own Default agent for quick actions.
• Select AI for transcription: chooses the AI connection used to transcribe recordings.
• Enable Transcribe: enables transcription with the selected AI connection.

The transcription connection must be available to the organization. If transcription is enabled, select a compatible AI connection before saving.

Default agents and transcription

These defaults serve different purposes:

Setting	Purpose and effect
Default LLM connection	Preferred model connection when preparing new agents and creating workspace agents. It does not choose a general-purpose agent.
Organization Default agent	The organization-level default agent for applicable entry points; user-specific defaults can be separate.
Default agent for quick actions	Agent used by desktop Quick Actions. It must belong to this organization and have its Desktop App interface enabled.
Workspace default agent	Agent created from the fixed system workspace template. It starts workspace chat, not automatic task execution. See Workspaces.

Choose the default model connection

In Feature availability → Agents, select Default LLM connection, review the other settings, and select Save. This requires Organization.ManageSettings. The connection must be a non-deleted model connection in this organization whose type supports agents, usable by you, and allowed by organization connection governance. Each agent still needs a model supported by its selected connection.

The creation paths resolve connections differently:

• New agent form: uses the available organization default. With no usable selected default, a single available model connection can be preselected; with several choices the user selects one. Explicit initial selections are preserved.
• Agent from template: prefers that default only when it supports the template model, otherwise tries another available compatible model connection. Check the prepared form before submitting.
• Workspace creation/recovery: uses the configured organization default, or an available organization system-default model connection if none is configured. A configured but incompatible or inaccessible connection causes creation to fail; it is not silently replaced by another connection.

Changing the organization default does not rewrite existing agents' connections, models, or conversation settings. It does not grant access to a connection. See agent Configuration and Library.

Quick Actions when an organization is created

New-organization provisioning creates a Quick Actions Agent from its system template when that template exists, and sets it as Default agent for quick actions. The agent is enabled for the Desktop App; Mobile App and Browser Extension are disabled. Organization members receive Can Use and the Admins team receives Can Edit. If the template is missing, that provisioning step is skipped. This behavior does not establish that existing organizations were backfilled.

For an existing organization, review the selected Quick Actions agent and its Desktop App availability. Enabling Quick actions and choosing its agent are separate from the model connection default. Actual desktop availability still depends on the client and its distribution.

Manual task execution uses the selected agent, then the task's assigned or source agent. Configure a Task created workflow for automatic processing.

Configure recording transcription and live transcription separately. The live transcription (Whisper) controls include its connection, model name, automatic language detection, and a language field when automatic detection is off. Leave the model name empty to use the provider default. For a fixed language, enter an ISO 639-1 code such as en or cs. Recording transcription has its own model and language settings.

To save organization settings, the user needs Organization.ManageSettings. Enabling a feature makes it available at organization level; roles and resource access still control who can use it.

Apps & Extensions

Use Apps & Extensions to control whether organization members can access:

• the Mobile App,
• the Browser Extension,
• the Desktop App,
• the Siesta AI MCP Server.

Disabling an item removes that distribution surface for members; it does not change the permissions, tools, or data available to an already configured agent.

When the MCP Server is enabled, members can copy its client configuration from Profile > Apps & Extensions. Administrators should review MCP access governance, and users connecting a compatible client should follow the MCP OAuth setup.

Content Import

Open Organization → Settings → Content import → Start import. Importing and checking packages require Content.Import; access to organization settings also follows its own permissions. This flow targets the current organization, not system templates shared across tenants. Imported skills and templates become usable by the organization; they do not grant access to agents, connections, or credentials.

The Content import section is at the bottom of Settings. Select Start import to open the import dialog.

Choose a source

Source	Action
Store	Select Connect to load the available package from the configured Store. Loading is not importing.
JSON	Paste the package JSON. The displayed tab is JSON, not Code.
File	Select or drop a JSON package file.

The dialog starts on Store. Select Connect to load the available package and continue to its review.

Alternatively, paste a package into JSON:

To load a package from disk, use File:

The package contains categories, skills, and agent templates. Use a package supplied by an approved maintainer. The source/file limit is 10 MiB; the normalized request sent for validation or import must fit 1 MiB, so a file accepted by the picker can still be too large to submit. These are separate limits.

Review and check changes

1. Select Review changes. This runs a dry run, comparing the package with organization content without importing it.
2. Review Categories, Agents (agent templates), and Skills. Search by name or description, filter by dry-run status, and select which items to include.
3. Review each template's skill dependencies. Removing a skill here removes that binding from the template selection; it does not delete the skill itself or automatically exclude its separate package item.
4. After changing selected items or bindings, select Check changes again. Check needed means the previous result no longer applies to the current selection.
5. Proceed only with a current successful check and at least one selected item. Failed items, an outdated check, or an empty selection prevent final import.

In Agents, review templates and their skill dependencies. This example shows a template and its associated skill:

In Skills, choose the individual skills to import. The status shows what the last check predicts; the skill is not saved until you select Import.

After changing a selection or template dependency, run Check changes again:

Status	Meaning in review	Meaning after a successful import
Created	The item would be created.	The item was created.
Updated	The matching item would change.	The item was updated.
Unchanged	No content change is planned.	No content change was required.
Failed	Validation or dependency resolution failed; inspect its messages.	Inspect the failure and verify actual state before retrying.

A review count is a prediction, not proof of a saved import. If the backend plan contains failures, it is not applied. Removing one item may invalidate references from another; recheck the complete selection rather than assuming dependencies were also removed.

Import and verify the result

Import performs the write for the checked selection. Review the result summary and each item's messages, then confirm the expected entries in Library → Organization Templates, Skills, and the relevant categories. Check actual state after a network failure before retrying. Do not assume that the import provides a one-click rollback.

Template synchronization documented for Library editing and Override existing template must not be assumed for imports. After importing an updated template, independently verify any existing linked agents before relying on a changed prompt.

Import errors

Problem	What to do
Enter valid JSON or invalid package	Correct JSON syntax and the package structure before review.
Check needed	Check the changed selection again.
Failed item or missing skill/category/model reference	Read its messages, repair the package or selection, and recheck.
Store cannot load	Check Store availability and sign-in with the deployment administrator; do not substitute an arbitrary URL.
Source or request is too large	Reduce the package; the 10 MiB source and 1 MiB normalized-request limits are independent.
Import is unavailable	Check Content.Import, organization settings access, a nonempty selection, and a current successful dry run.

Troubleshooting

If Settings cannot be saved:

• confirm that the selected default agent still exists and belongs to the organization,
• confirm that transcription is not enabled without a compatible AI connection,
• check feature availability before looking for a missing module elsewhere,
• verify that the import package uses the expected JSON structure,
• confirm that the intended reusable content appears in Library > Organization Templates after a successful import,
• ask the package owner or deployment administrator for the supported import behavior if the result is incomplete or unexpected.