For agent authors
Build agents that integrate with Multipliers
You don’t need to host on our platform to integrate. Any agent that emits a Multipliers Value Envelope (MVE) outcome — via MCP tool result, direct HTTP POST, or A2A protocol — shows up in the same value-chain rollup, the same governance dashboard, and the same per-run audit trail as a Multipliers-native agent.
This quickstart walks through the three integration paths most third-party authors choose: MVE-via-MCP (for Anthropic MCP servers), direct outcome POST (for Salesforce Apex Invocables, Slack bots, internal microservices), and A2A registration (for agents already speaking the Agent-to-Agent protocol).
What the full guide will cover
- MVE envelope schema — outcome kinds, amounts, idempotency, source attribution
- Path A: emit MVE on MCP tool result via _meta.multipliers_value
- Path B: POST /api/v1/outcomes directly (Tier B path, no MCP required)
- Path C: A2A AgentCard registration for cross-host agent federation
- Spec-cards — declarative policy contracts your agent operates under
- Idempotency keys, retry semantics, and at-least-once delivery
- Tenant API token scopes and how to request them from your tenant admin
In the meantime — read the integration spec
The full public integration spec v1.0 is already live. It covers every endpoint, every schema, and a 5-step quickstart.