MCP Integration

The fyrn MCP server lets AI agents manage your integrations through natural conversation. Agents can list flows, create new integrations, deploy, monitor, and troubleshoot — all through the Model Context Protocol.

What MCP is

The Model Context Protocol (MCP) is an open standard for connecting AI models to external tools and data sources. fyrn’s MCP server exposes 41 tools that give AI agents structured access to the platform.

Setting up the fyrn MCP server

Claude Desktop configuration

Add to your Claude Desktop config file (~/Library/Application Support/Claude/claude_desktop_config.json on macOS):

{
  "mcpServers": {
    "fyrn": {
      "command": "npx",
      "args": ["@fyrn/mcp-server"],
      "env": {
        "FYRN_API_URL": "https://api.fyrn.ai",
        "FYRN_API_KEY": "sk-your-agent-key"
      }
    }
  }
}

Agent API keys

Create an agent API key in the fyrn web UI under Settings → Agent Keys. Each key has:

Permission tier (1, 2, or 3) — controls which tools are available
Rate limit — per-second and daily quotas
Allowed tools — optional whitelist of specific tools

Permission tiers

Tier	Label	What agents can do
1	Use	List flows, search, inspect status, trigger flows, view schedules, test adapters, view lookups
2	Build	Everything in tier 1, plus: create instances, create flows, deploy, create adapters, create lookups, start batches
3	Manage	Everything in tier 2, plus: edit flows, pause/resume, retry messages, manage healing, update schedules, delete lookups

Approval gate

Tools that create or modify resources (tier 2-3) require human approval:

Agent calls the tool → server returns approval_required with an action ID
Human reviews and approves in the web UI
Agent re-calls with approval_id → action executes

This ensures agents can’t make destructive changes without oversight.

All available tools

Tier 1 — Use

Tool	Description
`list_flows`	List all flows with status and health
`search_flows`	Search flows by name or description
`get_flow_status`	Detailed flow status with metrics
`trigger_flow`	Send a payload through a flow
`list_published_apis`	List published API endpoints
`call_published_api`	Call a published API endpoint
`list_schedules`	List scheduled flows
`get_batch_status`	Get batch job progress
`list_batch_jobs`	List recent batch jobs
`list_adapters`	List custom adapters
`test_adapter`	Test an adapter with sample input
`list_lookups`	List lookup tables
`get_lookup`	Get a lookup table with entries
`export_batch_errors`	Export batch job errors
`get_api_spec`	Get OpenAPI spec for a published API
`get_messages`	View message logs (requires tier 3 at runtime)

Tier 2 — Build

Tool	Approval	Description
`list_catalog`	No	Browse connector catalog
`create_instance`	Yes	Create connector instance
`test_instance`	No	Test connector connection
`create_flow`	Yes	AI-generate and create a flow
`preview_flow`	No	Preview transform without delivering
`clone_flow`	Yes	Clone a flow to another environment
`deploy_flow`	Yes	Deploy a draft flow
`trigger_schedule`	Yes	Manually trigger a schedule
`start_batch`	Yes	Start a batch job
`create_adapter`	Yes	Create a custom adapter
`create_lookup`	Yes	Create a lookup table
`set_lookup_entry`	Yes	Add/update lookup entry
`import_lookup_csv`	Yes	Bulk import lookup entries
`retry_batch_failed_records`	Yes	Retry failed batch records
`create_published_api`	Yes	Create a published API

Tier 3 — Manage

Tool	Approval	Description
`edit_flow`	Yes	Edit flow YAML, name, description
`pause_flow`	No	Pause an active flow
`resume_flow`	No	Resume a paused flow
`retry_message`	No	Retry a failed message
`list_healing_events`	No	View self-healing events
`update_schedule`	Yes	Update cron, timezone, enabled
`update_adapter`	Yes	Update adapter code
`delete_lookup`	Yes	Delete a lookup table
`delete_lookup_entry`	Yes	Remove a lookup entry
`approve_healing`	Yes	Approve a healing fix

See MCP Tools Reference for complete parameter details and JSON examples.

Example conversation

Here’s how an agent might build and deploy a flow:

User: Create a flow that syncs new Shopify orders to our NetSuite ERP

Agent: I’ll create that integration. Let me generate the flow… calls create_flow with description

Agent: I’ve generated a flow that maps Shopify order fields to NetSuite sales orders. Here’s the YAML — would you like me to deploy it?

User: Yes, deploy it

Agent: calls deploy_flow — requires approval

Agent: The deployment needs your approval. Please review and approve in the fyrn dashboard.

(Human approves)

Agent: calls deploy_flow with approval_id — deployed!

Agent: Your flow is live. Webhook URL: https://api.fyrn.dev/webhooks/receive/abc123. New Shopify orders will automatically sync to NetSuite.

Security considerations

PII filtering: Agent responses automatically redact sensitive config fields and PII
Rate limiting: Per-second and daily quotas per agent key
Loop detection: Repeated identical calls are blocked
Audit logging: Every tool call is recorded with agent key, input, output, and duration
Sandbox mode: Optional sandbox mode blocks deploy_flow entirely

What’s next

MCP Tools Reference — Complete parameter and example reference
CLI Usage — CLI alternative for the same operations
Creating Flows — Build flows with AI or manually