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

# Actions library

The full menu of what a workflow can do: set status, priority and assignment, start approvals, send messages, add tags, call webhooks, and act in your connected third-party tools

### Actions: what workflows can do

Mix native request actions, notifications, webhooks, and third‑party integrations. Here’s the menu you’ll use most.

**AI actions**

* Summarize thread, AI routing and Translate
* Let AI decide the priority, status, inbox, tags, service or applications, or write the quick reply — see [AI Actions](https://docs.siit.io/workflow/ai-actions.md)

**Request**

* Set request status
* Set priority
* Assign to (team inbox or individual) — **Assign to admin** offers **Specific admin**, one named admin, or **Rotate between admins**, a list of at least two admins taken in turn
* Associate to service
* Start approval(s)
* Send quick reply (write the message in the action, with variables; optionally attach files or link a knowledge article)
* Add a note (internal)
* Add tags
* Add followers
* Set resolution and close
* Snooze / Pause or resume SLA
* Create a group (create a named group holding 1 to 20 new requests in one step)

**Communication**

* Send a communication (email or portal notification)
* Post a message in a Slack channel (for swarms, incidents)
* DM requester or assignee via Slack/Teams

**Paths**

* Paths for conditional branches

**Webhook**

* Call external webhook with request payload

**Third-party applications**

* Identity and access (Okta, Azure AD, Google Workspace, JumpCloud—varies by connection)
  * Activate/suspend user
  * Reset password
  * Add/remove from application
  * Add/remove from group
  * Revoke/restore sessions
  * De-provision user
  * **Set Temporary Password** (Google Workspace, Microsoft Entra, JumpCloud) — pick **Generated by Siit** for a new random password on every run, or **Typed here** for a fixed one stored encrypted.
  * **Issue Temporary Access Pass** (Microsoft Entra) — a single-use pass that signs the requester in once so they can register Microsoft Authenticator or a passkey. Request triggers only.
  * **Send Backup Verification Code** (Google Workspace) — one backup verification code, which replaces every backup code the requester already had. Request triggers only.
  * **Invalidate Backup Verification Codes** (Google Workspace) — stops those codes working, once the requester has enrolled a new device.
  * Hand the requester a temporary password, an Okta temporary access code, a Microsoft Entra Temporary Access Pass, a Google backup verification code, or a computer's recovery key. These steps deliver the value to the requester themselves, so you don't add a message action after them — see [Secret codes](https://docs.siit.io/request-management/secret-codes.md).
* Device management (Jamf, Iru, JumpCloud, Intune) — request triggers only
  * **Send FileVault Recovery Key** — reads the key the MDM escrowed for the requester's Mac and sends it to them privately.
  * **Send Recovery Lock Password** — same, for the Recovery Lock password of an Apple silicon Mac (Jamf and Iru).
  * **Clear Passcode** (Iru, JumpCloud) and **Reset Passcode** (Intune) — remove the passcode from a phone or tablet so its holder can unlock it and set a new one. Nothing is sent to anyone: the MDM returns no passcode. The device has to be one the requester holds, so the step fails on a computer or on someone else's device.
  * **Clear Activation Lock** (JumpCloud) — clears the Activation Lock on an Apple device the requester holds, so the device can be set up again with another Apple Account.
  * **Lock Device** (Jamf, Iru, JumpCloud) — locks one device remotely. On a computer the step also handles the 6-digit unlock PIN; phones and tablets are locked without one. **Unlock PIN** decides where that PIN goes: **Send to the requester**, which only runs on a device the requester holds, or **Post as an internal note**, which keeps it for admins who can access the request. Either way it stays revealable for 24 hours — see [Secret codes](https://docs.siit.io/request-management/secret-codes.md). Jamf also takes an optional **Lock message**, with variables, shown on the locked screen.
  * **Unlock Account** (Jamf, Iru) — unlocks a local account on a computer after too many wrong passwords. **Account to unlock** is either **Device's account**, the single account the MDM reports, or **Custom**, a username you write with variables. The step fails when the device has no account or several, and lists them.
  * **Add Device To Groups** and **Remove Device From Groups** (Jamf) — change one device's membership of Jamf static groups. Pick a **Group type**, **Computer groups** or **Mobile device groups**, then the **Groups** themselves; the step only runs on a device of that kind. Groups are changed one by one in the order you list them, a group the device is already in or out of counts as done, and the step stops at the first group Jamf rejects, keeping the changes already made. Only static groups are offered: Jamf manages smart group membership itself.
  * **Restart Device** and **Shut Down Device** (Iru, JumpCloud) — send the command to one device managed by that MDM. Unlike the steps above, these two run on any device in the fleet, not only one the requester holds.
  * JumpCloud's device commands reach Apple devices enrolled in JumpCloud MDM only. A step whose MDM isn't connected never shows in the action picker.
  * **Run on** tells the step where to look for the device: **Requester's equipment**, the devices the requester holds, or **Equipment on the request**, the answers to the request's **Equipment** form questions. The two recovery steps default to **Requester's equipment**, which is what they have always used; the device commands above default to **Equipment on the request**. Every one of these steps needs exactly one device to be left once Siit has kept the ones it can run on, so it fails before anything is sent when none or several qualify. A recovery key is still only sent when that computer is assigned to the requester.
  * Each step also returns what it ran on — **Equipment**, **Equipment category**, **Equipment serial number**, **Equipment lifecycle**, **Equipment model** and **Equipment office location** — for later steps and branch conditions. See [Variables & Secrets](https://docs.siit.io/workflow/variables-and-secrets.md).
  * Intune's BitLocker key is not available to workflows — see [Secret codes](https://docs.siit.io/request-management/secret-codes.md).
* Issue trackers
  * Jira: create issue; sync status/links
  * Jira Service Management: create linked ticket
  * Linear: create issue; sync status/links
* More integrations
  * Add additional tools as you connect them; actions appear automatically in the list.
  * The actions you declare on a [custom integration](https://docs.siit.io/integrations/custom-integrations.md) appear here too, once they carry a workflow input definition.

The actions a role can pick from depend on the integrations that role is granted under **External actions** — see [Roles & permissions](https://docs.siit.io/workspace/roles-and-permissions.md). A step that uses an integration your role has since lost stays in the workflow and keeps running; you simply can't save a change to it, and it no longer shows in the action picker or the history filters.

### Action slugs

Every action carries a slug, a short lowercase name that identifies it inside its workflow. When nobody chose one, Siit generates it from the action's own identifier and the workflow editor hides it. Only a slug someone chose shows in grey under the action's name, so you can tell two steps of the same kind apart at a glance without reading a string of random characters on every card.

Slugs are how later actions reference an earlier action's output, as `{{actions.<slug>.<output>}}`. You set one through the API, or Assist does when it builds a workflow for you, and it survives a workflow being rebuilt, which the internal action identifiers do not. A slug can be renamed, which rewrites every reference to it, but it can no longer be cleared. See [Variables & Secrets](https://docs.siit.io/workflow/variables-and-secrets.md) for how to use an action's output further down a workflow.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.siit.io/workflow/actions-library.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
