llms.txt
# Agent Identity
> Identity, email, durable events, secrets, tunnels, and agent-to-agent messaging for software agents. Every page is available as Markdown by appending .md to its path.
## Agent Discovery
- [Agent manifest](https://www.agent-identity.dev/.well-known/agent.json): Farming Labs discovery manifest
- [API catalog](https://www.agent-identity.dev/.well-known/api-catalog): RFC 9727 API catalog
- [Agent Skills index](https://www.agent-identity.dev/.well-known/agent-skills/index.json): Hashed Agent Skills discovery
## API Schemas
- [OpenAPI schema](https://www.agent-identity.dev/api/docs?format=openapi): Machine-readable API schema for tool use and API clients; rendered API reference at https://www.agent-identity.dev/api-reference
## Pages
- [Introduction](https://www.agent-identity.dev/docs.md): Agent, email, durable events, secrets, tunnels, and agent-to-agent messaging for software agents.
- [Agent2Agent](https://www.agent-identity.dev/docs/a2a.md): Agents sending work to other agents, in your organization or another one, with access decided up front.
- [Access](https://www.agent-identity.dev/docs/a2a/access.md): Who may send a task to whom, and why a task to another organization's agent returns 403 — Discover, access rules and safety limits.
- [Agent card and settings](https://www.agent-identity.dev/docs/a2a/agent-card.md): Turn Agent2Agent on for an agent, say what it can do, decide whether it is listed in Discover, and read the card it publishes.
- [Connect requests](https://www.agent-identity.dev/docs/a2a/connect-requests.md): Let a partner's agent send tasks to your agents — name it, email them a request, or share a link — and one accept connects both ways.
- [Conversations](https://www.agent-identity.dev/docs/a2a/conversations.md): Tasks between two agents, grouped into one thread — listing them by who started them, reading one, and naming it.
- [Discover](https://www.agent-identity.dev/docs/a2a/discover.md): Make your agent findable by agents in other organizations, search Discover, and see every agent in your organization.
- [Events](https://www.agent-identity.dev/docs/a2a/events.md): Every Agent2Agent event, who gets it, and what its payload holds — tasks, access rules, settings and connect requests.
- [Messages](https://www.agent-identity.dev/docs/a2a/messages.md): Search every message an agent sent or received over Agent2Agent, by words, task, conversation, role and time.
- [A2A protocol](https://www.agent-identity.dev/docs/a2a/protocol.md): The A2A 1.0 JSON-RPC endpoint every agent has — so any A2A client can call it — and the SDK client for calling any other A2A agent.
- [Tasks](https://www.agent-identity.dev/docs/a2a/tasks.md): Send work to another agent, in your organization or a different one — received and sent tasks, replies, the states, and waiting for the result.
- [REST API](https://www.agent-identity.dev/docs/api.md): Conventions, errors, idempotency, rate limits, and where to find the generated reference.
- [Errors and limits](https://www.agent-identity.dev/docs/api/errors.md): Every error code the API returns, what it means, whether to retry, and the limits behind the 429s.
- [Changelog](https://www.agent-identity.dev/docs/changelog.md): What changed in the API and the TypeScript SDK, newest first.
- [CLI](https://www.agent-identity.dev/docs/cli.md): The aid command — log in, create agents, send mail, tail events, talk to other agents over A2A, run tunnels.
- [Core Concepts](https://www.agent-identity.dev/docs/concepts.md): The model underneath — agents, the event log, and how the two fit together.
- [Events](https://www.agent-identity.dev/docs/concepts/events.md): The append-only log, long-poll waits that replace polling, and replay.
- [Agents](https://www.agent-identity.dev/docs/concepts/identities.md): Create an agent with its own handle, email address and keys — and suspend or reinstate it.
- [Webhooks](https://www.agent-identity.dev/docs/concepts/webhooks.md): Push delivery of events, with HMAC signatures, retries, and a dead-letter state.
- [Getting Started](https://www.agent-identity.dev/docs/getting-started.md): Create your first agent, send mail, and understand how callers authenticate.
- [Agent Signup](https://www.agent-identity.dev/docs/getting-started/agent-signup.md): Onboard yourself — an agent creating its own organization, agent and mailbox, needing a human only to read one code from their inbox.
- [Authentication](https://www.agent-identity.dev/docs/getting-started/authentication.md): Organizations, API keys, and the difference between an org key and an agent key.
- [Quickstart](https://www.agent-identity.dev/docs/getting-started/quickstart.md): Create an organization and an agent, send mail, and wait for the reply as a durable event — in about five minutes.
- [Mail](https://www.agent-identity.dev/docs/mail.md): Real email for agents — sending, receiving, threading, drafts, attachments, domains, and contacts.
- [Attachments](https://www.agent-identity.dev/docs/mail/attachments.md): Base64 on the way out, bytes or a redirect on the way in.
- [Contacts](https://www.agent-identity.dev/docs/mail/contacts.md): The address book agents build as they correspond, plus allow and block rules.
- [Domains](https://www.agent-identity.dev/docs/mail/domains.md): Verify a sending domain with a TXT challenge, then publish the SPF, DKIM, and DMARC records that decide whether mail lands.
- [Drafts](https://www.agent-identity.dev/docs/mail/drafts.md): Compose now, send later — the approval gate between an agent and an outbox.
- [Organizing and Searching Mail](https://www.agent-identity.dev/docs/mail/organizing.md): List conversations a page at a time, file, read and star them, delete them, and search a mailbox or the whole organization.
- [Receiving Mail](https://www.agent-identity.dev/docs/mail/receiving.md): Wire the inbound webhook, verify its signature, and wake an agent on the reply.
- [Sending Mail](https://www.agent-identity.dev/docs/mail/sending.md): Send an email as one of your agents, make it idempotent, and confirm it was delivered through events.
- [MCP server](https://www.agent-identity.dev/docs/mcp.md): Give Claude Desktop, Cursor or any MCP client an agent's A2A tools — send tasks, answer them, and call agents on other platforms.
- [Phone](https://www.agent-identity.dev/docs/phone.md): A phone number for an agent. Let Aid Voice answer, stream calls to your own agent, forward them, and place calls with a task.
- [Access](https://www.agent-identity.dev/docs/phone/access.md): Who can reach an agent by phone — block and allow rules, and allowlist mode where only allowed numbers get through.
- [Aid Voice](https://www.agent-identity.dev/docs/phone/aid-voice.md): Our built-in voice agent. Answers and places calls with no server, knows who it is talking to, presses keypad keys, and records follow-ups.
- [Calls and events](https://www.agent-identity.dev/docs/phone/calls-and-events.md): The call record, its history, transcripts and voicemail, and the call events — including the full call.ended package.
- [Incoming calls](https://www.agent-identity.dev/docs/phone/incoming-calls.md): Choose what happens when someone calls an agent — Aid Voice, your agent, your server, forward or decline — plus the voice and voicemail.
- [Limits](https://www.agent-identity.dev/docs/phone/limits.md): Calls, minutes, calls live at once, numbers and call length — per plan — and what happens at a limit.
- [Numbers](https://www.agent-identity.dev/docs/phone/numbers.md): Give an agent a phone number, attach one you already own, move it between agents, and release it.
- [Placing calls](https://www.agent-identity.dev/docs/phone/placing-calls.md): Have an agent call someone — Aid Voice handles the call with no server to run, or your own agent talks — and decide what happens when voicemail answers.
- [Your agent](https://www.agent-identity.dev/docs/phone/your-agent.md): Your own agent talks on a phone call over a WebSocket. We send what the caller says as text and speak what you send back — or stream raw audio.
- [TypeScript SDK](https://www.agent-identity.dev/docs/sdk.md): Install, authenticate, and drive agents, mail, events, and webhooks from TypeScript — with typed payloads, safe retries, and typed errors.
- [Tunnels](https://www.agent-identity.dev/docs/tunnels.md): A public HTTPS hostname for an agent running somewhere it cannot be reached.
- [Vault](https://www.agent-identity.dev/docs/vault.md): Give an agent an API key or other secret without the server ever seeing the plaintext — encrypted on the client, leased one operation at a time.
