Get your keys.

Starter tier — 1,000 manifests per month on monthly delivery cadence. No credit card required. Your agent can self-provision via the API, or enter your email below and we'll send your credentials.

For Agents

Self-provision

Your AI agent can generate credentials programmatically. No human signup required. Run this command or have your agent call the endpoint directly.

curl -s https://synorb.com/connect

Returns key, secret, MCP token, Streamable HTTP Core MCP config, legacy SSE URL, and Advanced connector URL. Works with MCP-compatible agents and partner connector hosts.

Response

{ "api_key": "sk_...", "api_secret": "ss_...", "mcp_token": "mt_...", "core_connector_url": "https://mcp.synorb.com/mcp", "core_auth": "Authorization: Bearer mt_...", "core_sse_url": "https://mcp.synorb.com/sse?token=mt_...", "advanced_connector_url": "https://mcp.synorb.com/advanced/mcp", "advanced_sse_url": "https://mcp.synorb.com/advanced/sse?token=mt_...", "mcp_server": { ... } }

For Humans

Email my keys

Enter your email and we'll generate free credentials and send them to your inbox. Same access — just delivered via email.

What you get

REST API Authenticate with key + secret headers.
Key Your public identifier. Include as api-key header in API requests.
Secret Your private key. Include as secret header. Keep this safe — shown only once.
MCP SERVERS Use Core by default; use Advanced only for configured workflows.
Token Authentication token for MCP. Use as Authorization: Bearer; query-token URLs remain available for older connector hosts.
Core URL Five-tool Core default over Streamable HTTP: profile, stream-search, catalog, details, manifests.
Advanced URL Full tool surface for account, digests, sync ontology, labels, prompts, and low-level content tools.

The Starter tier includes 1,000 manifests per month with MCP + API access on monthly delivery cadence. Need more volume, webhooks, or S3 delivery?

View plans and pricing →

Provisioning limits: Starter credentials are rate-limited per IP, per email, and per browser fingerprint to prevent abuse. New tokens have a per-token MCP rate-limit floor of 30 requests/minute on the Starter tier; paid plans scale this up. Validation requests (profile reads) have their own separate limit and do not consume Manifest quota. Excess requests return HTTP 429 with a Retry-After header.