firecrawl config with hidden token

[lilac quest]

v2026.5.20

How do I configure the firecrawl plugin to:

1. use the cloud firecrawl site
2. to use a token stored in .env (FIRECRAWL_API_KEY)

[rich anvil]

To help others find answers, you can mark your question as solved via Right click solution message -> Apps -> ✅ Mark Solution

[lavish tartan]

For v2026.5.20, use the plugin-owned config path. Firecrawl cloud is already the default (https://api.firecrawl.dev), but you can set it explicitly.

Put the key in the gateway env, usually:

# ~/.openclaw/.env
FIRECRAWL_API_KEY=fc-your-key-here

Then config, for Firecrawl as web_fetch fallback / scrape:

{
  tools: {
    web: {
      fetch: {

[lavish tartan]

        enabled: true,
        provider: "firecrawl",
      },
    },
  },
  plugins: {
    entries: {
      firecrawl: {
        enabled: true,
        config: {
          webFetch: {
            apiKey: { source: "env", provider: "default", id: "FIRECRAWL_API_KEY" },
            baseUrl: "https://api.firecrawl.dev",
          },
        },

[lavish tartan]

      },
    },
  },
}

If you also want Firecrawl as web_search:

{
  tools: {
    web: {
      search: {
        enabled: true,
        provider: "firecrawl",

[lavish tartan]

      },
    },
  },
  plugins: {
    entries: {
      firecrawl: {
        enabled: true,
        config: {
          webSearch: {
            apiKey: { source: "env", provider: "default", id: "FIRECRAWL_API_KEY" },
            baseUrl: "https://api.firecrawl.dev",
          },
        },
      },
    },

[lavish tartan]

  },
}

You can also omit apiKey entirely and OpenClaw will fall back to FIRECRAWL_API_KEY, but I prefer the SecretRef form above because it makes the dependency explicit without storing the token in config.

After editing ~/.openclaw/.env, restart the gateway so the daemon picks it up. Docs: https://docs.openclaw.ai/tools/firecrawl and https://docs.openclaw.ai/tools/web-fetch

[lilac quest]

I only want to use it for web_fetch fallback. What does that config look like?

[lavish tartan]

For only web_fetch fallback, keep it to this:

{
  tools: {
    web: {
      fetch: {
        enabled: true,
        provider: "firecrawl",
      },
    },
  },
  plugins: {
    entries: {
      firecrawl: {
        enabled: true,

[lavish tartan]

        config: {
          webFetch: {
            apiKey: { source: "env", provider: "default", id: "FIRECRAWL_API_KEY" },
            baseUrl: "https://api.firecrawl.dev",
            onlyMainContent: true,
            maxAgeMs: 172800000,
            timeoutSeconds: 60,
          },
        },
      },
    },
  },
}

And in ~/.openclaw/.env:

[lavish tartan]

FIRECRAWL_API_KEY=fc-your-key-here

Then restart the gateway. You do not need tools.web.search or plugins.entries.firecrawl.config.webSearch for this use case.

Tiny note: baseUrl is optional because that cloud URL is the default, and apiKey can also be omitted because FIRECRAWL_API_KEY is a fallback. I’d still keep the SecretRef line because it documents exactly where the hidden token comes from.

[lilac quest]

where are "onlyMainContent", "maxAgeMs" and "timeoutSeconds" documented?

[lavish tartan]

They’re in two places:

• Firecrawl tool docs, under “Configure Firecrawl scrape + web_fetch fallback”: https://docs.openclaw.ai/tools/firecrawl
• Web fetch docs, in the Firecrawl fallback config example: https://docs.openclaw.ai/tools/web-fetch

The generated config reference also lists the defaults under plugins.entries.firecrawl.config.webFetch: https://docs.openclaw.ai/gateway/configuration-reference

Current meanings/defaults:

• onlyMainContent: extract only main page content, default true
• maxAgeMs: max Firecrawl cache age in milliseconds, default 172800000 / 2 days
• timeoutSeconds: Firecrawl scrape request timeout, default 60

So for a minimal config you can leave those three out and just keep the provider + env SecretRef.