The CloudRadial AutomationAI Firecrawl extension lets an agent read the web. Firecrawl loads a page in a real browser and returns clean markdown an agent can work with, or maps a site to list its pages. This article covers what the extension can do, the API key and secrets it needs, and the skills that ship with it. It is for workspace users who build automations (Admin or Owner Role).
- What the Extension Covers
- Getting the API Key
- Providing the Credentials to Your Runner
- Read-Only Tools
- Skills
- Tips for Reading Web Pages
What the Extension Covers
The extension calls the Firecrawl API and provides two tools:
| Tool | What it does |
| Read a page | Loads one URL in a real browser and returns its content as markdown by default, with navigation, headers, and footers stripped. It can also return HTML, the page's links, or a screenshot. |
| Map a site | Lists the URLs Firecrawl can find on a site, optionally narrowed by a search term, so the agent can choose which page to read |
Use it when an agent needs the content of a specific page, such as an article, a documentation page, release notes, or a status page. Install it from Extensions like any other catalog extension. It appears in the catalog as Firecrawl. See Browsing and installing extensions for the general install flow.
Getting the API Key
The extension authenticates with a Firecrawl API key, which starts with fc-. Sign in to the Firecrawl dashboard and open the API Keys page, where you can create, view, and revoke your team's API keys. Your account always keeps at least one active key, so to replace a key, create the new one before you revoke the old one.
You can also point the extension at a self-hosted Firecrawl instance. It uses the same API, so only the base URL changes.
Providing the Credentials to Your Runner
Store two secrets in your runner's Key Vault with these exact names:
| Secret | Value |
Firecrawl-ApiUrl |
The Firecrawl API base URL, https://api.firecrawl.dev/v2, or the base URL of your self-hosted instance |
Firecrawl-ApiKey |
The Firecrawl API key |
Store the key exactly as Firecrawl issued it; the extension sends it as a Bearer token for you. For how to add secrets, see Accessing and updating your runner's Key Vault.
The agent reads the credentials from the runner and never asks you for them. Once the secrets are in place, ask the AI Playground to read a public page you know to confirm the connection works.
Read-Only Tools
The extension is read-only. Both tools only fetch and extract page content: they can't click, type, submit forms, or otherwise change a page, so neither requires approval.
Skills
The extension ships two skills, both read-only:
| Skill | Use it to |
| Read a Web Page | Get the content of a specific page, such as an article, a documentation page, release notes, or a status page, as clean markdown |
| Discover a Site's Pages | Find which pages exist on a site before reading one, for example to find the pricing page on a vendor's site |
Tips for Reading Web Pages
- For a normal read, just give the URL. Markdown of the page's main content is the default and the cleanest input for an agent.
- If you don't know the exact page, ask the agent to map the site first, then read the page it finds
- To capture a screenshot, ask for it together with the markdown. Firecrawl returns the screenshot as a URL or encoded image in the result.
- The extension can't fill in a form or click through a page, so it reads only what loads at the URL
If you are still having trouble, we’re here to help! Submit a ticket here for assistance, and don’t forget to check our status page to ensure there are no outages in your area.
Comments
0 comments
Please sign in to leave a comment.