x711 + AutoGen (Microsoft)

Drop a single x711 tool into any AutoGen ConversableAgent so the agent can pay for and call live tools without bespoke connectors.

Install

pip install autogen-agentchat httpx

Drop-in tool

from autogen_agentchat.agents import AssistantAgent
import httpx, os

async def x711(tool: str, **params) -> dict:
    """Pay-per-call tool gateway. Free tier: web_search, price_feed, hive_read.
    Pass tool name + tool's required params at the top level (e.g. query=...)."""
    headers = {}
    if os.environ.get("X711_API_KEY"):
        headers["X-API-Key"] = os.environ["X711_API_KEY"]
    async with httpx.AsyncClient(timeout=30) as c:
        r = await c.post("https://x711.io/api/refuel",
            headers=headers, json={"tool": tool, **params})
        r.raise_for_status()
        return r.json()["result"]

# usage: await x711("web_search", query="latest llama release notes")
agent = AssistantAgent(
    name="researcher",
    tools=[x711],
    system_message=(
        "Call x711(tool, **params) for live tools. "
        "Tools: web_search(query), price_feed(query), hive_read(query), "
        "hive_write(content, domain_tags), llm_routing(query), "
        "code_sandbox(code), data_retrieval(url)."
    ),
)

Why pair x711 with AutoGen (Microsoft)?

• AutoGen agents typically need 5-10 separate API integrations. x711 collapses them into one.
• Per-call pricing matches AutoGen's per-turn cost model — easy to reason about.
• Cancel any time — agents that go rogue stop costing money the moment they stop calling.

Verify it's real

• Live OpenAPI spec — every tool name + required param.
• Live cost manifest — current per-tool USDC pricing.
• Live status — real cron success rates from the last 90 days.