Feature

A clean /api/v1 surface, built for programmatic sending.

Bring leads in and send through the same programmatic surface: POST /api/v1/leads to register a lead, then a send endpoint that runs the lead through suppression and the warmed pool. The full contract is documented as an OpenAPI 3.1 spec at /api/v1/docs.

RESTMCP
rest-api
curl -X POST https://api.cognilead.ai/api/v1/sends \
  -H "Authorization: Bearer pk_live_..." \
  -H "Content-Type: application/json" \
  -d '{ "to": "lead@acme.com", "subject": "...", "body_md": "..." }'

The problem

What makes this hard today.

01

Most sequencer APIs are an afterthought

UI-first tools often bolt an API onto a product designed for a human clicking through screens, leaving it thinner than an infrastructure-grade surface.

02

No stable contract to build against

Without an OpenAPI spec, integrating against a moving target is risky for anything you plan to run in production.

03

Auth and scoping matter for delegated access

Teams managing multiple clients or sub-accounts need real key scoping, not one shared credential for everything.

How CogniLead helps

The mechanism, step by step.

  1. OpenAPI 3.1 spec, published at /api/v1/docs

    The full contract — endpoints, request/response shapes, auth — is documented and versioned, not inferred from trial and error.

  2. Bearer pk_live_ keys with RBAC

    Scope-limited API keys per tenant, with default-denied customer sub-keys for delegated (reseller/agency) access.

  3. The same surface an MCP agent uses

    The REST API and the MCP server (/features/mcp-server) sit on top of the same underlying pipeline — nothing is exposed to one surface and hidden from the other.

FAQ

Questions about REST API.

Where is the full API reference?+

At /api/v1/docs — the published OpenAPI 3.1 spec with every endpoint, auth, and schema.

How is authentication scoped?+

Bearer pk_live_* keys, scope-limited per tenant, with default-denied sub-keys available for delegated access (see /legal/security §5).

Does every send go through suppression automatically?+

Yes — suppression and reputation checks run server-side before dispatch on every call, regardless of which client is making it.

Related reading

REST API — CogniLead