Built for the agentic era

Agents runtime
for real-world
applications.

Deploy agents with built-in memory, tools, sandboxing, and orchestration. From one agent to thousands.

Start building Read the docs

Works with every coding agent

terminal|agent session

The full stack for AI agents

Everything to run agents in production

Serverless Agents

Define once, run from any app

Your agents live in the cloud. Every app — mobile, web, CLI, backend — calls the same endpoint. No infra to manage.

OpenAI-compatible API

Sub 10 seconds to POST /v1/chat/completions.

Built-in agent memory

Persistent memory per agent. Context survives across sessions.

Built-in tools

Agents ship with 20+ tools out of the box.

Sandboxed execution

Each task runs in an isolated sandbox with real system access.

Auto-scale

Sandbox pool scales automatically with load.

soon

Generative UI

Agents stream structured UI components back to your app.

Agent Teams

beta

Multi-agent orchestration

Agents work together. The orchestrator dispatches tasks, manages dependencies, and coordinates across your entire team.

Orchestration engine

Automatic task dispatch, result collection, and coordination.

Task dependencies

Define execution order. Tasks wait for their dependencies.

Missions

Group tasks into missions with deterministic workflow.

Concurrent agents

Multiple agents execute in parallel across sandboxes.

Long-running tasks

Agents can run for minutes or hours.

Scheduling

Trigger agents on a recurring schedule.

LLM as Judge

beta

Automated quality assessment

Every task is assessed against your expectations. Multiple reviewers score the work, a judge arbitrates, and the agent retries with targeted feedback.

LLM Judge evaluation

Task results can be assessed by an LLM reviewer that scores quality automatically.

G-Eval scoring

Define custom rubrics and weighted dimensions. Scores from 1 to 5.

Fix & retry pipeline

Failed tasks are automatically retried with targeted feedback until they pass.

Framework agnostic

Works with your stack

Polpo exposes an OpenAI-compatible API. Any language, any framework, any HTTP client.

How it works

Three steps to production

Define your agents

Define agents in a single JSON file — role, model, tools, skills, identity. Version-controlled, portable.

json

[{
  "name": "coder",
  "role": "Senior Engineer",
  "model": "anthropic:claude-sonnet-4-5",
  "systemPrompt": "Write clean, tested TypeScript...",
  "allowedTools": ["bash", "read", "write", "edit"],
  "skills": ["frontend-design", "testing"],
  "reasoning": "medium",
  "reportsTo": "tech-lead",
  "maxConcurrency": 3,
  "identity": {
    "displayName": "Marco Rossi",
    "title": "Full-Stack Developer",
    "tone": "Technical but concise"
  }
}]

Deploy with CLI or API

One CLI command or one API call. Polpo provisions isolated infrastructure and gives you a live endpoint.

$ polpo-cloud deploy --dir ./my-project

Deploy from ./my-project/.polpo

Core:
  Teams ........... yes
  Agents .......... yes
  Memory .......... yes
  Skills .......... yes
  Vault ........... yes
  Sessions ........ yes

Deployed: 3 team(s), 8 agent(s), 2 memory file(s),
          4 skill(s), 12 vault entry(ies)

Call from anywhere

OpenAI-compatible API. Your mobile app, web app, CLI, backend — anything that speaks HTTP talks to your agents.

Agent-first

Your coding agent
already knows Polpo

Any coding agent that can run CLI commands or call APIs can create, configure, and deploy Polpo agents in seconds. One prompt. Done.

Polpo comes with skills — install from skills.sh.

Click to copy

terminal

> Create an agent for customer support and deploy it

Write(.polpo/agents.json)

+ agent: "support-bot"

+ model: "anthropic:claude-sonnet-4-5"

+ tools: ["read", "http_fetch", "email_send"]

Bash(polpo deploy)

Deploying agents... 3.4s

Your agent is live at

https://api.polpo.sh/v1 · agent: "support-bot"

Use cases

What teams build with Polpo

App builders

Build AI-powered apps with agents as the backend. Ship features, not infrastructure.

like Lovable, Bolt, Replit Agent

AI employees

Agents with identity, memory, and skills that work as autonomous team members. 24/7, no burnout.

Zero-human companies, digital workforce

Voice agents

Plug voice agents into your call center or assistant. Real tools, real actions, real-time.

Support bots, sales assistants

Agent control panels

Build dashboards to manage your agents. Monitor runs, inspect logs, control execution — all via API.

Internal tools, admin panels

Autonomous workflows

Multi-agent missions that run on schedule. Triage tickets, generate reports, monitor systems.

DevOps, support, data pipelines

Integrate in your coding agent

Any coding agent can create and deploy Polpo agents via CLI or API in seconds. One command from your IDE.

Claude Code, Codex, Cursor, Windsurf

Stop managing infrastructure.
Start shipping agents.

Free tier included. No credit card. Deploy your first agent in under a minute.

Start building for free

npm install -g polpo-ai

Agents runtimefor real-worldapplications.