# Site Watch

Site Watch accepts a public website address, retrieves the page, summarises the response, and retains a history of checks. The starting template is useful for inspecting a page over time without connecting an external account.

## Check a page

1. Create **Site Watch** and open the app.
2. Enter the complete HTTP or HTTPS address of the public page you want to inspect.
3. Choose **Check now**.
4. Follow the workflow status and inspect the resulting history row.

The history includes the URL, status, summary, and any note supplied by processing. Compare successive checks to understand what the retrieved content says over time. Keep the URL consistent when you want to follow the same page; a redirect or a different path can change the material retrieved.

## Understand retrieval limits

The workflow sees the response available to its retrieval step. A browser session may show additional content behind sign-in, scripts, location choices, or consent prompts. A successful response can still be a challenge page or a generic error message. Inspect the returned status and content when a summary does not match what you see in your browser.

The template’s summary is not a guaranteed exact change comparison. Small edits can be missed or paraphrased differently. If you need a precise comparison, adapt the workflow to retain the relevant text and calculate a deliberate difference between two identified checks. Do not infer that an unchanged summary proves the source page is unchanged.

## Repeat or automate

**Check now** performs an explicit check. To run periodically, configure a [scheduled workflow trigger](/docs/workflows/scheduled-triggers), choose the intended timezone behaviour, and inspect the first scheduled runs. Set a frequency appropriate to the source and your processing budget.

Add notification or destination steps only when the app should send an alert or update another service. A history row alone does not establish that a notification was delivered. Use run output and delivery status to distinguish collection from downstream action.

If a check is interrupted, inspect the history and run before repeating it. Preserve relevant results independently if you need a lasting record beyond the app’s normal lifecycle.

<!-- catalogue-reference -->

## Parameter reference

These are the starting template’s declared parameters. Your saved app can have its own version of the schema. A default applies when no explicit value is supplied; a required field still needs a valid value.

| Setting | Accepted value | Required | Default | What it controls |
| --- | --- | --- | --- | --- |
| title (`title`) | string; minimum characters 1; maximum characters 200 | Yes | Not specified | The app title; also woven into the provisioned workflow names. |


- `title`: exposed as an app interface parameter.

## Connection slots

| Slot | Required | Initially enabled |
| --- | --- | --- |
| No external or managed slot required by the starter | — | — |

Choose compatible saved connections for user-supplied slots. Managed slots are supplied by Strive. Disabling an optional slot stops its use without necessarily deleting the saved binding. See [connector policy](/docs/apps/connector-policy).

## Included workflows

| Workflow | Steps in the starting recipe | Trigger |
| --- | --- | --- |
| `watch` | 6 | manual |

Inspect the saved app’s workflows for changes made after creation. A workflow’s presence does not mean that every interaction runs every workflow.

## Output format: site_check

One check of a web page: the URL, its title, ok / blocked (refused by the SSRF guard) / error, the model's summary of what the page shows, and the failure note.

| Setting | Accepted value | Required | Default | What it controls |
| --- | --- | --- | --- | --- |
| url (`url`) | string | Yes | Not specified | — |
| title (`title`) | string | Yes | Not specified | — |
| status (`status`) | "ok" / "blocked" / "error" | Yes | Not specified | — |
| summary (`summary`) | string | Yes | Not specified | — |
| note (`note`) | string | Yes | Not specified | — |