@telaro/crewai

CrewAI tool descriptors for Telaro. Exposes the trust API as @tool functions any CrewAI Agent can call.

Install

npm install @telaro/crewai

Usage

The TS package emits the descriptors; pair with a small Python wrapper that turns them into CrewAI @tools:

// generate-tools.ts (run at build time)
import { crewToolDescriptors } from "@telaro/crewai";

console.log(JSON.stringify(
  crewToolDescriptors({ apiBase: "https://telaro.xyz" }),
  null, 2,
));

# crew.py
from crewai import Agent, Task, Crew
from crewai.tools import tool
import json, requests

@tool("check_agent_policy")
def check_policy(pubkey: str, min_bond: int = 100_000_000, min_score: int = 700) -> dict:
    """Check if a Telaro agent meets a (min_bond, min_score) policy."""
    res = requests.get(f"https://telaro.xyz/api/agent/{pubkey}").json()
    a = res["agent"]
    return {
        "passes": a["bond_atomic"] >= str(min_bond) and a["score"] >= min_score and not a["frozen"],
        "score": a["score"], "bond_atomic": a["bond_atomic"], "frozen": a["frozen"],
    }

risk_analyst = Agent(role="Risk Analyst", tools=[check_policy])

Why this exists

CrewAI’s strength is multi-agent orchestration. Pairing a "Risk Analyst" CrewAI agent with Telaro lets you build crews that vet on-chain agents programmatically — useful for any DApp delegating to multiple bonded agents at once.

See also

• Quickstart
• @telaro/sdk