# Ravi

> Ravi gives AI agents their own identity — phone numbers, email addresses, and encrypted credential storage — so they can sign up for services, receive verification codes, and operate autonomously on the web.

## Getting Started

- [Introduction](https://docs.ravi.app/getting-started/introduction.txt): What is Ravi and why your AI agents need their own identity.
- [Installation](https://docs.ravi.app/getting-started/installation.txt): Install the Ravi CLI and authenticate your account.
- [Sign up for a service in 5 minutes](https://docs.ravi.app/getting-started/quickstart.txt): Create an agent Identity and use it to sign up for an external service, receive verification codes, and store credentials — end to end.
- [Authentication](https://docs.ravi.app/getting-started/authentication.txt): How authentication works in Ravi — OAuth login, token management, and encryption setup.
- [Troubleshooting & FAQ](https://docs.ravi.app/getting-started/troubleshooting.txt): Solutions to the most common errors developers hit when integrating Ravi.
- [Build a Ravi Integration](https://docs.ravi.app/getting-started/build-an-integration.txt): Tool-agnostic guide to building a Ravi integration from scratch — provision identities, handle verification flows, store credentials, and recover from errors.

## Core Concepts

- [Identities](https://docs.ravi.app/core-concepts/identities.txt): Create, manage, and switch between agent Identities.
- [Email](https://docs.ravi.app/core-concepts/email.txt): Send and receive email from your agent's dedicated address.
- [Phone & SMS](https://docs.ravi.app/core-concepts/phone-and-sms.txt): Receive SMS messages and verification codes on your agent's dedicated phone number.
- [Credential Vault](https://docs.ravi.app/core-concepts/credential-vault.txt): E2E-encrypted password and secret storage for agent credentials.
- [E2E Encryption](https://docs.ravi.app/core-concepts/encryption.txt): How Ravi's zero-knowledge encryption protects your agent's credentials.

## CLI Reference

- [CLI Overview](https://docs.ravi.app/cli/overview.txt): Architecture and design of the Ravi CLI.
- [Command Reference](https://docs.ravi.app/cli/commands.txt): Complete reference for all Ravi CLI commands.

## Integrations

- [Claude Code](https://docs.ravi.app/integrations/claude-code.txt): Use Ravi with Claude Code for autonomous agent identity management — as a direct user or as an orchestrator spawning parallel subagents.
- [OpenClaw](https://docs.ravi.app/integrations/openclaw.txt): Use Ravi as an OpenClaw plugin for real-time email, SMS, and credential management.
- [Agent Skills](https://docs.ravi.app/integrations/agent-skills.txt): Skill files that teach any AI agent how to use Ravi.
- [API Authentication](https://docs.ravi.app/integrations/api-keys-and-auth.txt): How API authentication works — JWTs, Identity scoping, and billing.
- [Symphony (OpenAI)](https://docs.ravi.app/integrations/symphony.txt): Use Ravi to supply credentials, agent identities, and OTP verification inside Symphony autonomous coding runs.
- [Claude Code Bridge](https://docs.ravi.app/integrations/claude-code-bridge.txt): Use Ravi with claude_code_bridge to give multi-model agent teams persistent, externally-routable identities.
- [Aperant](https://docs.ravi.app/integrations/aperant.txt): Add persistent agent identity to Aperant's parallel AI coding sessions — credential vault, email verification, and per-agent authentication via Ravi.
- [cmux](https://docs.ravi.app/integrations/cmux.txt): Use Ravi with cmux to give each parallel coding-agent workspace a real identity — provisioned email, phone, and credentials — from the moment the session starts.
- [Ariana](https://docs.ravi.app/integrations/ariana.txt): Give every Ariana VPS agent a real identity — provisioned email, phone, and encrypted credential vault via Ravi — so autonomous cloud agents can sign up, verify, and authenticate without human interruption.
- [Automaker](https://docs.ravi.app/integrations/automaker.txt): Give every Automaker agent worktree a real identity — provisioned email, phone, and encrypted credential vault via Ravi — so autonomous development tasks complete end-to-end without breaking the loop.
- [Claude Squad](https://docs.ravi.app/integrations/claude-squad.txt): Run parallel AI coding agents with per-session Ravi identities for isolated email, phone, and credential access.
- [t3code](https://docs.ravi.app/integrations/t3code.txt): Integrate Ravi with t3code for per-thread credential injection, autonomous agent identities, and traceable commits in your AI coding workflows.
- [Constellagent](https://docs.ravi.app/integrations/constellagent.txt): Run parallel AI coding agents with per-session identity isolation using Ravi.
- [vibe-kanban](https://docs.ravi.app/integrations/vibe-kanban.txt): Integrate Ravi with vibe-kanban to give coding agents real identities per workspace — provisioned email, phone, and secrets vault for agents that do more than write code.
- [vibe-tree](https://docs.ravi.app/integrations/vibe-tree.txt): Run Claude agents in parallel git worktrees with per-worktree Ravi identities for credential isolation and notification routing.
- [Agent Deck](https://docs.ravi.app/integrations/agent-deck.txt): Terminal mission control for AI coding agents, with MCP pooling and conductor orchestration.
- [assistant (kcosr)](https://docs.ravi.app/integrations/assistant-kcosr.txt): Panel-based personal assistant with plugin architecture for multi-agent workflows
- [Accomplish](https://docs.ravi.app/integrations/accomplish.txt): Give your Accomplish AI desktop agent its own identity — isolated API keys, provisioned email, and phone verification for autonomous workflows.
- [Agent Orchestrator (ComposioHQ)](https://docs.ravi.app/integrations/agent-orchestrator.txt): Run fleets of parallel AI coding agents with Ravi-powered credential injection, per-agent identity, and human-in-the-loop approval via email or SMS.
- [AI Maestro](https://docs.ravi.app/integrations/ai-maestro.txt): Multi-machine agent orchestration dashboard with AMP messaging — give each agent a real Ravi identity for external-world operations.
- [amux + Ravi](https://docs.ravi.app/integrations/amux.txt): Run parallel coding agents with isolated identities. amux manages the terminals; Ravi manages the credentials.
- [CodexMonitor + Ravi](https://docs.ravi.app/integrations/codexmonitor.txt): Run multiple Codex agents across isolated workspaces with persistent, secure identity via Ravi.
- [dmux + Ravi](https://docs.ravi.app/integrations/dmux.txt): Run parallel AI coding agents with isolated identities — one Ravi credential bundle per worktree, automatically provisioned and revoked.
- [Dorothy + Ravi](https://docs.ravi.app/integrations/dorothy.txt): Add Ravi as Dorothy's 6th MCP server — credential-free agent operations, real-world identity for your Super Agent, and per-agent secret isolation across your entire parallel fleet.
- [emdash + Ravi](https://docs.ravi.app/integrations/emdash.txt): Give every parallel agent in emdash its own real identity — centralized API key management, SSH credential provisioning, per-agent secrets isolation, and autonomous service sign-ups across 22 providers.
- [HumanLayer / CodeLayer + Ravi](https://docs.ravi.app/integrations/humanlayer.txt): Give every parallel coding agent its own verified identity for human-in-the-loop approval flows — email-based approvals from a real agent address, per-session isolated inboxes, and a full audit trail in standard email threads.
- [JAT + Ravi](https://docs.ravi.app/integrations/jat.txt): Give every agent in your JAT swarm its own real identity — email, phone, and credential vault — powered by Ravi.
- [Jean + Ravi](https://docs.ravi.app/integrations/jean.txt): Give every parallel AI coding session its own verified identity — scoped GitHub tokens, isolated credentials, and automatic lifecycle management for Jean's multi-worktree development environment.
- [lalph + Ravi](https://docs.ravi.app/integrations/lalph.txt): Give every issue-driven agent its own verified identity — encrypted credential vault, per-preset agent personas, and built-in verification flows for lalph's autonomous dev loop.
- [Mux + Ravi](https://docs.ravi.app/integrations/mux.txt): Give every parallel coding agent workspace its own scoped identity — isolated credentials, verified email/phone for service sign-ups, and TOTP support for remote SSH agents — powered by Ravi alongside Coder's Mux.
- [Nimbalyst + Ravi](https://docs.ravi.app/integrations/nimbalyst.txt): Give each parallel coding agent its own verified identity — email, phone, and credential vault — inside Nimbalyst's multi-session workspace.
- [OpenKanban + Ravi](https://docs.ravi.app/integrations/openkanban.txt): Give every OpenKanban ticket its own scoped identity — isolated API keys, real email and phone for service sign-ups, and TOTP support for 2FA checkpoints — powered by Ravi alongside OpenKanban's ticket-as-worktree model.
- [Subtask + Ravi](https://docs.ravi.app/integrations/subtask.txt): Give every subagent in a Subtask parallel workflow a real, scoped identity — credentials, disposable emails, phone numbers, and TOTP — provisioned automatically at task dispatch and cleaned up on merge.
- [Superset + Ravi](https://docs.ravi.app/integrations/superset.txt): Give every parallel coding agent its own real identity — email, phone, and credentials — so your worktree-based workflows can run fully autonomously.
- [Aizen](https://docs.ravi.app/integrations/aizen.txt): Give each parallel Aizen workspace its own Ravi identity — real email, phone, and credential vault scoped to the project.
- [babyagi3](https://docs.ravi.app/integrations/babyagi3.txt): Connect the minimal self-configuring personal agent to Ravi for email, SMS, and credential identity.
- [CashClaw](https://docs.ravi.app/integrations/cashclaw.txt): Autonomous agent that bids on work, executes tasks, collects payment, and self-improves — with Ravi providing persistent identity and secure credentials.
- [ClawWork](https://docs.ravi.app/integrations/clawwork.txt): Economic survival benchmark that turns AI models into real earners — Ravi provides isolated identity and credential management per competing agent.
- [DenchClaw](https://docs.ravi.app/integrations/denchclaw.txt): Fully managed OpenClaw framework for CRM automation and outreach agents, running locally on your Mac.
- [hermes-agent](https://docs.ravi.app/integrations/hermes-agent.txt): NousResearch's self-improving personal agent — cloud-native, multi-platform messaging, autonomous skill creation.
- [1code](https://docs.ravi.app/integrations/1code.txt): Ravi identity and secrets for parallel coding agents in the 1code orchestration platform.
- [VibeCraft](https://docs.ravi.app/integrations/vibecraft.txt): RTS-style infinite canvas workspace for running parallel AI coding agents, with Ravi providing per-agent identity and credential isolation.
- [Ravi MCP Server](https://docs.ravi.app/integrations/mcp.txt): Use Ravi as a Model Context Protocol server — expose identity, email, SMS, and vault tools to any MCP-compatible agent or orchestrator.
- [lemon (Elixir / BEAM)](https://docs.ravi.app/integrations/lemon.txt): Integrate Ravi with lemon and other BEAM-native agent runtimes using Elixir subprocess or direct REST API patterns.
- [LettaBot](https://docs.ravi.app/integrations/lettabot.txt): Wire Ravi into LettaBot for a persistent phone number, encrypted API key storage, and SMS OTP handling across Telegram, Signal, and voice channels.
- [LobsterAI](https://docs.ravi.app/integrations/lobsterai.txt): Use Ravi with LobsterAI to give each agent a real provisioned email, phone number, and credential vault — injectable as native Claude Agent SDK tool definitions.
- [picoclaw](https://docs.ravi.app/integrations/picoclaw.txt): Use Ravi with picoclaw on RISC-V, ARM, and other edge devices — headless auth, pre-provisioned identities, REST-only polling.
- [Supacode](https://docs.ravi.app/integrations/supacode.txt): Run 50+ parallel coding agents with shared Ravi identity — credentials, email, and secrets without the friction.
- [GhostClaw](https://docs.ravi.app/integrations/ghostclaw.txt): Give your always-on GhostClaw agent its own email address, phone number, and encrypted credential vault via Ravi.
- [Mercury](https://docs.ravi.app/integrations/mercury.txt): Use Ravi with Mercury to give each agent space a persistent email address, phone number, and credential vault — fully wired into Mercury's TypeScript extension system.
- [NanoClaw](https://docs.ravi.app/integrations/nanoclaw.txt): Use Ravi with NanoClaw to give each container-isolated agent a real email address, phone number, and credential vault — accessed via lightweight fetch() calls, no CLI subprocess required.

## Use Cases

- [Service Signup & Login](https://docs.ravi.app/use-cases/service-signup.txt): Complete workflows for signing up, logging in, and completing 2FA using Ravi.
- [Multi-Agent Setup](https://docs.ravi.app/use-cases/multi-agent-setup.txt): Run multiple agents with separate Identities for isolation and parallel operation.
- [Service Signup & Login with OpenClaw](https://docs.ravi.app/use-cases/openclaw-service-signup.txt): How OpenClaw agents autonomously sign up for services, verify accounts, log in, and manage credentials using Ravi identity tools.
- [Real-World Scenarios](https://docs.ravi.app/use-cases/real-world-scenarios.txt): Practical examples of agents using Ravi for autonomous web tasks.
- [Async Human Approval via Email](https://docs.ravi.app/use-cases/human-approval.txt): Let agents request human sign-off on high-stakes actions using their Ravi email identity — no webhook, no dashboard, no infrastructure required.
- [Human Approval in OpenClaw Agents](https://docs.ravi.app/use-cases/openclaw-human-approval.txt): A non-blocking, memory-driven human-in-the-loop pattern for OpenClaw agents — using Ravi email and SMS, skills, memory, and heartbeats instead of polling loops.
- [Agent-to-Agent Communication](https://docs.ravi.app/use-cases/agent-to-agent-email.txt): Route tasks between agents using Ravi email as a message bus — no broker, no queue, no shared state.
- [Production Patterns](https://docs.ravi.app/use-cases/production-patterns.txt): Python and TypeScript patterns for programmatic identity management, parallel agent isolation, and retry-based polling in production systems.
- [Harness Integration](https://docs.ravi.app/use-cases/harness-integration.txt): How orchestrators and CI harnesses create per-workspace Ravi identities at dispatch time without touching global state.
- [TypeScript REST API Integration](https://docs.ravi.app/use-cases/typescript-rest-api.txt): Use Ravi from TypeScript with direct REST API calls — no CLI subprocess required. Covers token loading, identity management, email/SMS polling, SSE streaming, and vault secrets.
- [Parallel Environments with Worktrees](https://docs.ravi.app/use-cases/parallel-environments.txt): Use .ravi/config.json per worktree root for automatic, zero-config identity isolation across parallel agents.

## API Reference

- [API Overview](https://docs.ravi.app/api/overview.txt): Architecture and conventions of the Ravi REST API.
- [API Endpoints](https://docs.ravi.app/api/endpoints.txt): Complete reference for the Ravi REST API endpoints.

## Security

- [Security Model](https://docs.ravi.app/security/security-model.txt): How Ravi protects your agent's identity, credentials, and communications.

## API

- [Error Reference](https://docs.ravi.app/api/error-reference.txt): Complete error codes, HTTP status codes, and recovery steps for Ravi CLI and REST API.