diff --git a/src/content/docs/reference/api-and-sdk/index.mdx b/src/content/docs/reference/api-and-sdk/index.mdx
index 64c09666..2d150990 100644
--- a/src/content/docs/reference/api-and-sdk/index.mdx
+++ b/src/content/docs/reference/api-and-sdk/index.mdx
@@ -93,7 +93,7 @@ See the [**Python SDK**](https://github.com/warpdotdev/oz-sdk-python) or [**Type
 
 ### Key Endpoints
 
-**The Agents API exposes three primary endpoints:**
+**The Agents API exposes these primary endpoints:**
 
 *   `POST /agent/run`
 
@@ -104,8 +104,14 @@ See the [**Python SDK**](https://github.com/warpdotdev/oz-sdk-python) or [**Type
 *   `GET /agent/runs/{runId}`
 
     Fetch full details for a single run, including session link and resolved configuration.
+*   `POST /agent/runs/{runId}/followups`
 
-All endpoint semantics, query parameters, and [error codes](/reference/api-and-sdk/troubleshooting/errors/) are documented on the [Agents API Reference](/api).
+    Send a follow-up message to an existing run, the same capability the Slack and Linear integrations use to continue a run. The server routes the message based on the run's current state (queued, running, or ended), so you can steer or continue a run programmatically. A `200` response means the follow-up was accepted; observe the updated state with `GET /agent/runs/{runId}`. The follow-up message and optional query `mode` are described by the `RunFollowupRequest` model.
+*   `POST /agent/runs/{runId}/cancel`
+
+    Cancel a run that is currently queued or in progress. Returns the ID of the cancelled run.
+
+The follow-up and cancel endpoints are also available in the SDKs (for example, `runs.submit_followup` and `runs.cancel`). All endpoint semantics, query parameters, and [error codes](/reference/api-and-sdk/troubleshooting/errors/) are documented on the [Agents API Reference](/api).
 
 ---
 
@@ -122,6 +128,7 @@ The API shares a set of reusable models across endpoints. Detailed JSON schemas,
 * `RunCreatorInfo`
 * `RunState`
 * `RunSourceType`
+* `RunFollowupRequest`
 * `AmbientAgentConfig`
 * `MCPServerConfig`
 * `Error`
diff --git a/src/content/docs/reference/api-and-sdk/quickstart.mdx b/src/content/docs/reference/api-and-sdk/quickstart.mdx
index 547b26ff..33e00fcd 100644
--- a/src/content/docs/reference/api-and-sdk/quickstart.mdx
+++ b/src/content/docs/reference/api-and-sdk/quickstart.mdx
@@ -91,6 +91,7 @@ You can also view and manage all runs in the [Oz dashboard](https://oz.warp.dev/
 
 ## Next steps
 
+* **Send a follow-up to a run** - Use `POST /agent/runs/{runId}/followups` (or `runs.submit_followup` in the SDKs) to send a follow-up message to a run, the same way the Slack and Linear integrations continue a run. The server routes the message based on the run's current state. See the [Oz API reference](/reference/api-and-sdk/) and the [Agents API Reference](/api).
 * **Read the full API reference** - [Oz API](/reference/api-and-sdk/) documents all endpoint parameters, query filters, and response schemas.
 * **Explore the SDKs** - [Python SDK](https://github.com/warpdotdev/oz-sdk-python) and [TypeScript SDK](https://github.com/warpdotdev/oz-sdk-typescript) include typed request/response models, retries, and error handling.
 * **See a real-world example** - [Demo: Sentry monitoring with SDK](/reference/api-and-sdk/demo-sentry-monitoring-with-sdk/) shows how to build a webhook handler that triggers agents from production errors.