Skip to main content
orchagent is an AI agent deployment platform. Build agents in Python, JavaScript, or plain English — deploy them to secure cloud sandboxes — and give your team a URL to run them via CLI, API, or web UI.

​
How It Works

​
For Developers (Building & Deploying)

1. Create agent: orch init my-agent
2. Define behavior (prompt + schema, code, or LLM loop)
3. Publish via CLI: orch publish
4. Agent is live — callable via CLI, API, or web UI
5. Team members run it from their workspace
6. See usage, run history, and logs in dashboard

​
For Teams (Running Agents)

1. Get access to your team's workspace
2. Run agents: orch run team/agent-name
3. Or open the web UI — fill a form, click Run
4. Results displayed, logged, and tracked

​
The Four Types

orchagent has four canonical types:
TypeWhat it doesExample
promptSingle LLM call with your prompt + schemaSentiment analyzer, summarizer, translator
toolYour Python/JS code runs in a sandboxSecurity scanner, PDF parser, Discord bot
agentLLM tool-use loop with custom toolsTest fixer, code reviewer, research agent
skillPassive knowledge (markdown) — not runnableCoding standards, security rules
Simple rule:
  • Want a single LLM answer? Use type: "prompt" (orch run)
  • Want your code to execute? Use type: "tool" (orch run)
  • Want the LLM to iterate with tools? Use type: "agent" (orch run)
  • Want to share knowledge? Use type: "skill" (orch skill install)
  • Want to DELEGATE to sub-agents? Export agents (orch install)
Executable types (prompt, tool, agent) are:
  • On-demand or always-on (run_mode) — each call is independent, or deploy as a service
  • Specific - one task, done well
  • Callable - simple API or CLI call; optionally callable by other agents
Skills use the industry-standard Agent Skills format (adopted by Anthropic, Google, Microsoft, OpenAI, Cursor) and can be installed locally or composed with agents.

​
Key Features

FeatureDescription
Secure sandboxesAgents run in isolated E2B sandboxes. No infra to manage.
CLI & APIDeploy and run agents from terminal or via HTTP API
orch-hq desktop appVisual command centre with terminal, canvas, messages, and tasks
Agent messagesAgents send you briefings, alerts, and notifications
Agent tasksAgents create tasks with due dates, priorities, and project tags
IntegrationsConnect Sentry, Slack, and more — agents use the credentials
Web Run UINon-developers run agents via a web form
Team workspacesOrganize agents by team, control access
Run historyEvery execution is logged with inputs, outputs, duration
Scheduled runsRun agents on a cron schedule or webhook trigger
Always-on servicesDeploy agents as persistent services (Discord bots, workers)
BYOK modelBring your own LLM API keys
Secrets vaultStore API keys securely in your workspace

​
BYOK: Bring Your Own Key

orchagent uses a Bring Your Own Key model. Developers provide their own LLM API keys (OpenAI, Anthropic, Gemini). Why BYOK:
  • Authors don’t pay for others’ usage
  • No middleman markup on LLM costs
  • Direct billing with your LLM provider
API keys are separate from subscriptions (Claude Pro/Max, ChatGPT Plus). You’ll need to set up API billing at console.anthropic.com or platform.openai.com.

​
Next Steps

Quickstart

Deploy your first agent in minutes

API Reference

Authentication, endpoints, errors

CLI Commands

orch login, run, publish

Agent Types

The four types — prompt, tool, agent, skill

orch-hq

Desktop command centre

Scheduling

Cron schedules and webhooks