Gazebo exposes an MCP (Model Context Protocol) server that AI tools can connect to for credential retrieval. Any MCP-compatible client — Cursor, Claude Code, Windsurf, or a custom setup — can use the same endpoint.
Endpoint
https://app.gazebohq.com/api/mcp
Authentication is via the agent bearer token passed as an Authorization header.
Available tools
get_identity
Returns the calling agent's name and full access policy. Call this first to confirm which services are accessible before fetching credentials. No credentials are returned; nothing is logged.
get_identity()
→ {
status: "ok",
agent_id: "...",
name: "cursor-dev",
accessible_services: [
{ service: "stripe", allowed_methods: ["GET", "POST"] },
{ service: "github", allowed_methods: ["GET"] }
]
}
get_credential
Retrieves the credential for a named service. Requires service and method — the method you declare you intend to use. Every call is logged.
See the get_credential reference for full documentation.
list_audit_events
Returns recent credential access events for the calling agent. Useful for verifying access, debugging denials, or auditing usage.
list_audit_events({ limit: 10 })
→ {
status: "ok",
count: 3,
events: [
{ service: "stripe", method: "POST", outcome: "success", timestamp: "..." },
{ service: "github", method: "GET", outcome: "denied", timestamp: "..." }
]
}
create_workflow
Creates a Gazebo workflow to configure an external service — e.g. set up Stripe webhooks, add a Cloudflare DNS record. The workflow is AI-planned, shown to the user in the Gazebo dashboard, and only executed after explicit approval.
Recommended usage pattern
1. get_identity() → confirm which services are accessible
2. get_credential(...) → fetch the credential you need
3. [use the credential] → make your API call
4. list_audit_events(...) → optionally verify the access was logged