Durable execution and guardrails for AI agents

agent.py

from langchain.agents import create_agentfrom langchain_openai import ChatOpenAIfrom rebuno import Agent, http_client, tool @tool("send_email", idempotency="at_most_once")async def send_email(body: str) -> dict:    """Email a written brief to a colleague."""    return await mail.send("[email protected]", body) async def process(query: str) -> dict:    llm = ChatOpenAI(model="gpt-5.5", http_async_client=http_client())    graph = create_agent(model=llm, tools=[send_email])    result = await graph.ainvoke({"messages": [{"role": "user", "content": query}]})    return {"answer": result["messages"][-1].content} agent = Agent("mailer")agent.run(process)

agent.py

from openai import AsyncOpenAIfrom pydantic_ai import Agent as PydanticAgentfrom pydantic_ai.models.openai import OpenAIResponsesModelfrom pydantic_ai.providers.openai import OpenAIProviderfrom rebuno import Agent, http_client, tool @tool("send_email", idempotency="at_most_once")async def send_email(body: str) -> dict:    """Email a written brief to a colleague."""    return await mail.send("[email protected]", body) async def process(query: str) -> dict:    client = AsyncOpenAI(http_client=http_client())    model = OpenAIResponsesModel("gpt-5.5", provider=OpenAIProvider(openai_client=client))    result = await PydanticAgent(model, tools=[send_email]).run(query)    return {"answer": result.output} agent = Agent("mailer")agent.run(process)

agent.py

from crewai import LLM, Agent as CrewAgent, Crew, Taskfrom crewai.tools import tool as crewai_toolfrom rebuno import Agent, execution, tool @crewai_tool("send_email")@tool("send_email", idempotency="at_most_once")async def send_email(body: str) -> dict:    """Email a written brief to a colleague."""    return await mail.send("[email protected]", body) async def process(query: str) -> dict:    # CrewAI builds its own HTTP client, so Rebuno intercepts at a gateway.    ctx = execution()    llm = LLM(        model="openai/gpt-5.5",        base_url=GATEWAY_URL,        extra_headers={            "rebuno-execution-id": ctx.id,            "rebuno-dispatch-id": ctx.dispatch_id,            "rebuno-dispatch-attempt": str(ctx.dispatch_attempt),            "rebuno-agent-id": ctx.agent_id,            "rebuno-agent-secret": SECRET,        },    )    writer = CrewAgent(role="writer", goal=query, llm=llm, tools=[send_email])    task = Task(description=query, expected_output="a short brief", agent=writer)    result = await Crew(agents=[writer], tasks=[task]).kickoff_async()    return {"answer": str(result)} agent = Agent("mailer")agent.run(process)

agent.ts

import { createOpenAI } from "@ai-sdk/openai";import { generateText, tool } from "ai";import { z } from "zod";import { Agent, defineTool, rebunoFetch } from "rebuno"; const sendEmail = defineTool({  name: "send_email",  idempotency: "at_most_once",  execute: async ({ body }: { body: string }) => mail.send("[email protected]", body),}); const emailTool = tool({  description: "Email a written brief to a colleague.",  inputSchema: z.object({ body: z.string() }),  execute: sendEmail,}); async function process(input: { query: string }) {  const openai = createOpenAI({ fetch: rebunoFetch });  const { text } = await generateText({    model: openai("gpt-5.5"),    prompt: input.query,    tools: { send_email: emailTool },  });  return { answer: text };} const agent = new Agent("mailer");await agent.serve({ port: 5000 }, process);

agent.ts

import { createOpenAI } from "@ai-sdk/openai";import { Agent as MastraAgent } from "@mastra/core/agent";import { createTool } from "@mastra/core/tools";import { z } from "zod";import { Agent, defineTool, rebunoFetch } from "rebuno"; const sendEmail = defineTool({  name: "send_email",  idempotency: "at_most_once",  execute: async ({ body }: { body: string }) => mail.send("[email protected]", body),}); const emailTool = createTool({  id: sendEmail.name,  description: "Email a written brief to a colleague.",  inputSchema: z.object({ body: z.string() }),  execute: sendEmail,}); async function process(input: { query: string }) {  const openai = createOpenAI({ fetch: rebunoFetch });  const assistant = new MastraAgent({    id: "mailer",     name: "Mailer",    instructions: "Email the brief when asked.",    model: openai("gpt-5.5"),    tools: { send_email: emailTool },  });  const { text } = await assistant.generate(input.query);  return { answer: text };} const agent = new Agent("mailer");await agent.serve({ port: 5000 }, process);

agent.ts

import { ChatOpenAI } from "@langchain/openai";import { tool } from "@langchain/core/tools";import { createAgent } from "langchain";import { z } from "zod";import { Agent, defineTool, rebunoFetch } from "rebuno"; const sendEmail = defineTool({  name: "send_email",  idempotency: "at_most_once",  execute: async ({ body }: { body: string }) => mail.send("[email protected]", body),}); const emailTool = tool(sendEmail, {  name: sendEmail.name,  description: "Email a written brief to a colleague.",  schema: z.object({ body: z.string() }),}); async function process(input: { query: string }) {  const model = new ChatOpenAI({ model: "gpt-5.5", configuration: { fetch: rebunoFetch } });  const graph = createAgent({ model, tools: [emailTool] });  const result = await graph.invoke({    messages: [{ role: "user", content: input.query }],  });  return { answer: result.messages.at(-1)?.content };} const agent = new Agent("mailer");await agent.serve({ port: 5000 }, process);