Poacher help
Use Poacher from an API or AI agent
Discover sources, read your account data, and optionally grant permission to rename or pause alerts.
Discover supported interfaces
The developer hub links to the OpenAPI schema, public source metadata and a public discovery MCP server. Source metadata explicitly requires exact-scope qualification; it is not a blanket support guarantee.
Read your own account data
Create a scoped, expiring API key under Developer access. Choose only the read permissions your integration needs. Keys can read the owning account’s permitted balance, alert or result data and cannot read another account. Keep bearer keys in server-side secret storage, never a public URL or browser-delivered code.
Current agent boundaries
The API can run an explicitly approved paid preview with work:execute permission, a customer-set daily allowance and a per-request ceiling. Actual provider expense is charged at 50% gross margin, including unsuccessful work. It cannot activate ongoing alerts or change billing settings; an approved manual check notifies only the alert’s existing connections. The public MCP endpoint provides discovery; a separate account MCP endpoint accepts scoped bearer keys for scoped account tools; the optional alerts:write permission can rename or pause an alert with revision and retry protection. Clients can paste a scoped key in a secret Authorization header, or sign in through OAuth 2.1 with PKCE (dynamic client registration and client ID metadata documents are supported) and approve the permissions in Poacher; the grant becomes a scoped, expiring key you can disconnect under Developer access. Use the app for monitoring changes and review the current developer documentation before integrating.