Fetch.ai Agentverse API

Fetch.ai has published a FastAPI route into Agentverse, giving developers a documented way to connect Python web apps to its agent network. (docs.agentverse.ai) The FastAPI guide shows a minimal agent with a `/status` health check and a `/chat` endpoint, built with FastAPI and `uagents_core`, then exposed to Agentverse through a public internet endpoint. (docs.agentverse.ai) Fetch.ai’s current documentation says Agentverse lets developers “launch and manage native or external agents,” connect them with ASI:One, and track performance, discovery, and ranking from one platform. (docs.agentverse.ai) Agentverse is not just a hosting panel. Fetch.ai describes it as a directory and growth layer where agents can be registered, searched, and surfaced to users and other agents across the broader ecosystem. (docs.agentverse.ai) That search layer already exposes a public API pattern. Fetch.ai’s search documentation shows developers can query `agentverse.ai/v1/search` with filters for agent state, category, type, and protocol, then get back fields including interaction counts, status, and timestamps. (innovationlab.fetch.ai) The company’s hosted-agent API also points to programmatic deployment. Its API reference includes endpoints to create a user agent with `POST /v1/hosting/agents` and to start a specific agent with `POST /v1/hosting/agents/:address/start`. (docs.agentverse.ai, docs.agentverse.ai) Fetch.ai says external agents need to speak the Agent Chat Protocol, a standard message format that lets agents receive requests and send structured replies across Agentverse and ASI:One. The same protocol is used in its guides for FastAPI, uAgents, and A2A, or agent-to-agent, adapters. (docs.agentverse.ai, docs.agentverse.ai, docs.agentverse.ai) Fetch.ai’s public product page says Agentverse is built to make agents discoverable, measure success rates, and tune performance, not just keep them running in the cloud. (fetch.ai) That framing helps explain the FastAPI push: instead of asking developers to rebuild agents in a new stack, Fetch.ai is trying to pull ordinary Python services into its own registry, search, and communication system. (docs.agentverse.ai, docs.agentverse.ai) The next test is whether developers treat Agentverse as infrastructure or as a marketplace. Fetch.ai now has documentation for both the plumbing and the discovery layer; adoption will depend on whether those registered agents actually attract traffic and interactions. (docs.agentverse.ai, innovationlab.fetch.ai)