Firecrawl and Dome
Scrape the pages you meant to, and refuse the rest. Firecrawl turns any URL into clean content for an agent, and searches, maps and crawls sites over a hosted MCP server. It also opens browser sessions and runs its own agent, more reach than most jobs need. Behind Dome, three tools stay open and scrapes stay on the sites you list.
Dome provides a governed gatewayfor Firecrawl.
Scrapes on listed sites
A rule reads the url of every scrape and map. Your supplier's docs and sec.gov are scraped, and any other site is refused.
Three tools open, the rest closed
Search, scrape and map stay open. Crawls, browser sessions, monitors and Firecrawl's own agent are closed until you open them.
No personal data in the request
A guard reads search queries and URLs alike. A card number, a Social Security number or an email address in either blocks the call.
Firecrawl behind the Gatewayin three steps.
Add Firecrawl's server, guard its arguments and deploy the rules. There's no catalog sync: the tools list as soon as the server is added.
Add the Firecrawl MCP server
One Firecrawl API key serves every agent. Dome sends it as a Bearer token, and the Gateway lists Firecrawl's tools once the server is added.
shelldome tools add --name firecrawl \ --url https://mcp.firecrawl.dev/v2/mcp \ --auth-method api-key \ --credential-type shared \ --authorization "Bearer $FIRECRAWL_API_KEY" \ --gateway prod-gatewayBlock personal data in arguments
The filter runs on requests. A match stops the call before Firecrawl fetches anything.
shellcat > query-leak.json <<'EOF'{"json":{"components":[{"fieldActions":[{ "matcher":{"contentMatchers":[{"ssn":{}},{"creditCard":{"requireLuhn":true}},{"email":{}}]}, "action":"FILTER_ACTION_BLOCK"}]}]}}EOFdome guards filters create query-leak --config-from query-leak.jsondome tools guards filters set firecrawl --direction request --filters query-leakApply the rules
The site list is scoped to filings-reader. Simulate a scrape on and off the list, then deploy.
shelldome rules apply firecrawl-sites.cedar --agent filings-reader --name firecrawl-sites
Scrape two sites,search the rest.
The permit opens search, scrape and map. Forbids close every other Firecrawl tool, refuse any search that names Project Halcyon and refuse a scrape or map outside two sites.
cedar
permit (principal, action == Dome::Action::"mcp:call", resource is Dome::MCPTool)when { resource.connection_name == "firecrawl" && ["firecrawl_search", "firecrawl_scrape", "firecrawl_map"].contains(resource.tool_name)};
forbid (principal, action == Dome::Action::"mcp:call", resource is Dome::MCPTool)when { resource.connection_name == "firecrawl" && !["firecrawl_search", "firecrawl_scrape", "firecrawl_map"].contains(resource.tool_name)};
forbid (principal, action == Dome::Action::"mcp:call", resource is Dome::MCPTool)when { resource.connection_name == "firecrawl" && resource.tool_name == "firecrawl_search" && resource has arguments && resource.arguments has query && resource.arguments.query like "*Halcyon*"};
forbid (principal, action == Dome::Action::"mcp:call", resource is Dome::MCPTool)when { resource.connection_name == "firecrawl" && ["firecrawl_scrape", "firecrawl_map"].contains(resource.tool_name) && !(resource has arguments && resource.arguments has url && (resource.arguments.url like "https://docs.northwind-supply.com/*" || resource.arguments.url like "https://www.sec.gov/*"))};One call,two outcomes.
Switch the caller or the argument and watch the same call decide differently, with every decision landing in audit.
Site
agent filings-reader · acting as s.patel
firecrawl/firecrawl_scrape(url: "https://www.sec.gov/Archives/edgar/data/320193/000032019325000079/aapl-20250927.htm", formats: ["markdown"])
- Agentfilings-reader is registered and active
- Callers.patel verified through Okta
- RuleScrapes on www.sec.gov are allowed
DecisionAllowed
Three agentson Firecrawl.
Each one reads the web as clean text. Two scrape, and only the sites they're given.
docs-indexer
Index a supplier's docs
Maps a supplier's documentation site each night and scrapes changed pages into the search index.
filings-reader
Read public filings
Scrapes annual reports on sec.gov and pulls the sections an analyst asks for, and is refused on any other site.
deal-scout
Scout acquisition targets
Searches for companies that fit a thesis and reads the results as clean text, and a query naming the deal is refused.
Bringing it alltogether.
Connecting Firecrawl to registered agents, models, and identity in Dome completes a governed agent application.
Acting for
Agent
Control point
GatewayEvery call decided and audited
RulesGuardsQuotas
Tools
FirecrawlAgent service
Models
Common questionsabout Firecrawl.
More of whatDome works with.
Model
Claude Fable
Fable 5.1 from Anthropic and Amazon Bedrock in one failover pool, open to one group and capped by quota.
Provider
Anthropic
The Claude API behind the Model Broker: the key held in Dome, every call authorized, metered and audited.
MCP server
GitHub
The GitHub MCP server behind the Tool Gateway, with rules that decide each call on its owner and repository.
Identity
Microsoft Entra ID
Entra access tokens verified on every agent call, so rules read app roles and audit names the person.
Runtime
LangGraph
LangGraph agents with their model calls on the Model Broker and their MCP tools on the Tool Gateway.
Client
Claude Code
Claude Code on a Dome Gateway with per-developer sign-in, rules on every tool call, and audit by name.
Talk with ourFDE team.
Or start free, and register your first agent in minutes. Connect it through a gateway, give it rules, and see every call it makes in one audit trail.
Talk with our FDE team
Our engineers design, build and run governed agents for the part of your business you want to change. You see what they deliver.
Agent services: web, research, payments, voice and sandboxes
Web actions, research, payments, phone calls and code execution are where agents reach past your walls. Put each service behind the Gateway, and every call is authorized, metered and audited.