# agenticweb.wiki

> Reference map of the agentic web's standards layer: nine lifecycle steps, one record per protocol, every claim sourced.

## Start here
- [What the agentic web is](https://agenticweb.wiki/agentic-web/index.md)
- [The map](https://agenticweb.wiki/map/index.md)
- [Methodology](https://agenticweb.wiki/methodology/index.md)
- [Resources](https://agenticweb.wiki/resources/index.md)
- [Instances](https://agenticweb.wiki/instances/index.md)
- [News](https://agenticweb.wiki/news/index.md)

## Data
- [All records JSON](https://agenticweb.wiki/data/standards.json)
- [CSV](https://agenticweb.wiki/data/standards.csv)
- [Probe rules](https://agenticweb.wiki/data/probes.json)
- [OpenAPI](https://agenticweb.wiki/openapi.json)
- [Every article as one Markdown file](https://agenticweb.wiki/llms-full.txt)
- [Markdown sitemap](https://agenticweb.wiki/sitemap.md)
- [AGENTS.md](https://agenticweb.wiki/AGENTS.md)
- [MCP server card](https://agenticweb.wiki/mcp/server-card)

## Categories
- [1 Discover](https://agenticweb.wiki/category/discover/index.md): What is here, and where?
- [2 Identify](https://agenticweb.wiki/category/identify/index.md): Who is this agent? Who is this site?
- [3 Permit](https://agenticweb.wiki/category/permit/index.md): What may the agent do here?
- [4 Read](https://agenticweb.wiki/category/read/index.md): Give me the content
- [5 Act](https://agenticweb.wiki/category/act/index.md): Do something for me
- [6 Coordinate](https://agenticweb.wiki/category/coordinate/index.md): Talk to another agent
- [7 Pay](https://agenticweb.wiki/category/pay/index.md): Settle the bill
- [8 Present](https://agenticweb.wiki/category/present/index.md): Show a human the result
- [9 Account](https://agenticweb.wiki/category/account/index.md): Prove what happened; be judged

## Protocols
- [A2UI (Agent to UI) Protocol](https://agenticweb.wiki/protocols/a2ui/index.md): How can an agent describe and stream a framework-neutral, declarative UI to a renderer?
- [Ad Context Protocol](https://agenticweb.wiki/protocols/adcp/index.md): How can an AI agent discover ad inventory, execute a media buy, sync creative, and report performance across many ad platforms through one protocol?
- [Agent Attribution Protocol](https://agenticweb.wiki/protocols/agent-attribution-protocol/index.md): When an AI agent recommends a product and a sale happens later, on another device or session, can anyone prove which agent (and which recommendation) actually drove it?
- [Agent Audit Trail](https://agenticweb.wiki/protocols/agent-audit-trail/index.md): Is there a standard, tamper-evident log format any autonomous AI agent can write its actions to, so a regulator, auditor or victim can later reconstruct exactly what it did and why?
- [Agent Authorization Grant](https://agenticweb.wiki/protocols/aauth/index.md): How can a conversational AI agent, reached over a channel without browser redirect (phone, SMS), obtain a properly scoped OAuth access token without impersonating the user?
- [Agent Authorization Profile](https://agenticweb.wiki/protocols/aap/index.md): How can OAuth 2.0/JWT claims carry structured agent identity, task context, operational constraints, delegation chains, and human-oversight requirements for machine-to-machine agent calls?
- [Agent Client Protocol](https://agenticweb.wiki/protocols/agent-client-protocol/index.md): How does a code editor invoke and converse with an interchangeable coding agent it did not build?
- [Agent Communication Protocol](https://agenticweb.wiki/protocols/agent-communication-protocol/index.md): How did agents, applications, and people exchange multimodal REST messages before merging into A2A?
- [Agent Connect Protocol (AGNTCY)](https://agenticweb.wiki/protocols/agent-connect-protocol/index.md): How does one agent (or an orchestrator) invoke and configure a remote agent it did not build, over a plain HTTP API?
- [Agent Directory Service](https://agenticweb.wiki/protocols/ads/index.md): Given a required skill, which agents provide it, where are they, and how can I fetch a verifiable, content-addressed record describing them?
- [Agent Discovery Protocol](https://agenticweb.wiki/protocols/adp/index.md): How does an agent escalate from a cheap DNS lookup, to a well-known JSON metadata document, to a real-time WebSocket session, using only as much machinery as the task needs?
- [Agent Identity and Discovery](https://agenticweb.wiki/protocols/aid/index.md): Given a domain, where is the agent and which protocol should I speak, with cryptographic proof the endpoint controls the matching key?
- [Agent Identity Protocol (AIP)](https://agenticweb.wiki/protocols/aip/index.md): How can an autonomous agent prove a decentralized identity and a verifiable chain of delegated authority from its human or organizational principal?
- [Agent Name Service v2 (ANS)](https://agenticweb.wiki/protocols/ans/index.md): What domain-anchored identity belongs to an autonomous agent, and at what assurance tier can that identity be trusted?
- [Agent Network Protocol](https://agenticweb.wiki/protocols/anp/index.md): How can decentralized agents from different platforms authenticate, discover, and message each other without a central registry?
- [Agent Payments Protocol](https://agenticweb.wiki/protocols/ap2/index.md): Did the user actually authorize what this agent is about to buy and pay for, and can the merchant prove it later?
- [Agent Protocol](https://agenticweb.wiki/protocols/agent-protocol-aief/index.md): How does a client or dev-tool start a task on an autonomous agent, step it, and fetch results, regardless of which framework built the agent?
- [Agent Readability Spec](https://agenticweb.wiki/protocols/vercel-agent-readability/index.md): How readable is this site's content for AI agents, and what specifically should be fixed?
- [Agent Registration and Discovery Protocol](https://agenticweb.wiki/protocols/ardp/index.md): How can an autonomous agent register a stable identity, advertise its capabilities and current endpoints with a TTL, and be discovered by others with minimal metadata leakage?
- [Agent Skills](https://agenticweb.wiki/protocols/anthropic-agent-skills-format/index.md): How can reusable, on-demand procedural instructions and bundled resources be packaged so any compatible agent can load and run them?
- [Agent Transfer Protocol](https://agenticweb.wiki/protocols/agtp/index.md): How can AI agent traffic be identified, authenticated, and governed as a distinct application-layer protocol rather than blending into ordinary HTTP?
- [agent-permissions.json](https://agenticweb.wiki/protocols/agent-permissions-json/index.md): Which page-level interactions (clicking, form submission, file upload, media control, script execution) may an agent perform while driving a web page's UI, as opposed to just fetching it?
- [Agent-User Interaction Protocol](https://agenticweb.wiki/protocols/ag-ui/index.md): How does an agent stream state, generative UI, messages, and tool activity into a user-facing application in real time?
- [Agent2Agent Protocol](https://agenticweb.wiki/protocols/a2a/index.md): How do two independent agents, built on different frameworks, discover each other and delegate a task to completion?
- [Agentic Advertising Management Protocols](https://agenticweb.wiki/protocols/aamp/index.md): How do a buyer's agent and a seller's agent discover each other's inventory, negotiate terms and hand off a transaction inside the existing ad-tech stack?
- [Agentic Commerce Protocol](https://agenticweb.wiki/protocols/agentic-commerce-protocol/index.md): How do a buyer's agent and a merchant complete a deterministic, programmatic checkout without the agent becoming the merchant of record?
- [Agentic Real Time Framework](https://agenticweb.wiki/protocols/artf/index.md): How does a host advertising platform invoke a third party's agent service, deployed as a container inside its own infrastructure, to mutate a bid request in real time?
- [Agentic Resource Discovery](https://agenticweb.wiki/protocols/ard/index.md): Where are an organization's discoverable agentic resources (MCP servers, A2A agents, skills, registries), and how can they be found by search rather than pre-installed one at a time?
- [agents.json](https://agenticweb.wiki/protocols/agents-json/index.md): How does an agent learn which multi-step API call sequences ("flows") a service supports, on top of its raw OpenAPI surface?
- [AGENTS.md](https://agenticweb.wiki/protocols/agents-md/index.md): What build, test, and convention instructions should a coding agent follow in this repository?
- [agents.txt](https://agenticweb.wiki/protocols/agents-txt/index.md): Which agent-facing protocols (payments, auth, MCP, skills, A2A) does this site speak, and where?
- [Agora (agent marketplace protocol)](https://agenticweb.wiki/protocols/agora-commerce-protocol/index.md): How can an AI agent discover, hire, and pay another AI agent for a paid job, with escrowed on-chain settlement?
- [Agora Protocol](https://agenticweb.wiki/protocols/agora-protocol/index.md): How can two heterogeneous LLM agents communicate efficiently, moving from expensive natural-language negotiation to a cheap, mutually agreed structured protocol?
- [AI Agent Authentication and Authorization (AIMS)](https://agenticweb.wiki/protocols/aiagent-auth/index.md): Is there an IETF-endorsed way to authenticate and authorize an AI agent, and has the WIMSE working group actually taken this on?
- [AI Catalog well-known file](https://agenticweb.wiki/protocols/ai-catalog-well-known-file/index.md): Which AI-capable artifacts and registries does this domain publish? (superseded question -- see ard)
- [AI Crawler User-Agent Tokens in robots.txt](https://agenticweb.wiki/protocols/ai-crawler-tokens-robots/index.md): Which named AI crawler or agent user-agent tokens may a website allow or disallow in robots.txt, and what does each token actually control (search indexing, AI training, or live user-triggered fetches)?
- [ai.txt](https://agenticweb.wiki/protocols/ai-txt/index.md): May AI systems train on, scrape, index, or cache this site's content, and under what per-agent or per-path conditions?
- [AIPREF Content-Usage Attachment](https://agenticweb.wiki/protocols/aipref-content-usage-attachment/index.md): How does a publisher attach its AI-use preferences to content served over HTTP, so an automated system can find them?
- [AIPREF Vocabulary](https://agenticweb.wiki/protocols/aipref-vocabulary/index.md): What controlled set of values expresses a publisher's AI-use preferences, independent of how they are attached to content?
- [AITP (Agent Interaction & Transaction Protocol)](https://agenticweb.wiki/protocols/aitp/index.md): How do two agents representing different, mutually untrusted people or businesses (e.g. a personal assistant and an airline's booking agent) talk to each other, exchange forms and payment requests, and not just plain text?
- [AP2 Checkout Mandate](https://agenticweb.wiki/protocols/ap2-checkout-mandate/index.md): What exactly did the user approve for checkout?
- [Apache Ossie (formerly Open Semantic Interchange)](https://agenticweb.wiki/protocols/osi/index.md): How can an AI agent or BI tool read a metric or data-model definition and get the same meaning another tool intended?
- [api-catalog: A Well-Known URI and Link Relation to Help Discovery of APIs](https://agenticweb.wiki/protocols/api-catalog-rfc9727/index.md): Given a publisher's domain, where is the machine-readable list of the APIs it exposes?
- [auth.md](https://agenticweb.wiki/protocols/auth-md/index.md): How does an AI agent register a new account for its user on a service, without a human filling out a sign-up form or an OAuth redirect flow existing to authorize it?
- [Content Signals Policy](https://agenticweb.wiki/protocols/cloudflare-content-signals-policy/index.md): May content fetched under a robots.txt allow rule also be used for search indexing, as live input to an AI answer, or for AI model training?
- [Coral Protocol (CoralOS)](https://agenticweb.wiki/protocols/coral-protocol/index.md): How do independently-built agents from a public marketplace get orchestrated together into one multi-agent system, with privacy isolation and built-in payment budgets, on top of MCP?
- [DNS for AI Discovery](https://agenticweb.wiki/protocols/dns-aid/index.md): Given a domain, where is its agent endpoint, what protocol does it speak, and how can that be verified via DNSSEC/DANE?
- [DNS-Based Agent Naming](https://agenticweb.wiki/protocols/dan/index.md): Given a domain, which AI agents does it publish, and how can I locate and cryptographically bind to their service endpoint using DNS alone?
- [EMV Agentic Payments – Framework for Specifications](https://agenticweb.wiki/protocols/emv-agentic-payments/index.md): When a card network, issuer, merchant, or wallet sees an AI-agent-initiated card payment, how do they consistently find out what the consumer actually authorised — recurring purchase, spending cap, post-transaction change — rather than each building a bespoke, one-off signal?
- [ERC-8004: Trustless Agents](https://agenticweb.wiki/protocols/erc-8004/index.md): Who is an agent, and what onchain identity, reputation and validation record does it carry?
- [ERC-8183: Agentic Commerce](https://agenticweb.wiki/protocols/erc-8183/index.md): How can two agents (a client and a provider) settle a job on-chain, with funds held in escrow and released only once an evaluator attests the work was done?
- [ERC-8226 Regulated Agent Mandate](https://agenticweb.wiki/protocols/erc-8226/index.md): What scoped, time-bounded, and financially capped authority may an on-chain agent exercise on a regulated asset holder's behalf?
- [HTTP Message Signatures](https://agenticweb.wiki/protocols/http-message-signatures/index.md): How can HTTP messages carry end-to-end component integrity and proof of possession, independent of transport-layer security?
- [Identity Assertion JWT Authorization Grant](https://agenticweb.wiki/protocols/identity-assertion-authz-grant/index.md): How can an app or AI agent, already SSO-federated to a user's enterprise identity provider, obtain scoped access to a third-party API or MCP server without a fresh consent screen at that third party?
- [L402](https://agenticweb.wiki/protocols/l402/index.md): How can a client pay for, and then authenticate to, a metered API resource over the Lightning Network with no accounts or database-backed sessions?
- [llms.txt](https://agenticweb.wiki/protocols/llms-txt/index.md): Where can an agent find a concise, curated map of a site's content, written for language models rather than browsers?
- [LMOS (Language Model Operating System)](https://agenticweb.wiki/protocols/lmos/index.md): How does an agent or tool from one organization describe itself and get discovered by agents from another organization, across network boundaries?
- [Machine Payments Protocol](https://agenticweb.wiki/protocols/mpp/index.md): How can a machine client (an agent, script, or another server) pay for an HTTP resource using any payment network, not just one?
- [MCP Apps](https://agenticweb.wiki/protocols/mcp-apps/index.md): How does an MCP tool deliver an interactive UI (chart, form, dashboard) that renders inline in a chat client?
- [Model Context Protocol](https://agenticweb.wiki/protocols/mcp/index.md): How does an AI application call tools and pull structured context, prompts, and data from an external server?
- [Natural Language Interaction Protocol](https://agenticweb.wiki/protocols/nlip/index.md): Can this agent or service exchange natural-language, multi-modal messages using a shared, vendor-neutral envelope instead of a bespoke API?
- [NLWeb](https://agenticweb.wiki/protocols/nlweb/index.md): How can a site expose a natural-language question-answering interface grounded in its own Schema.org-structured content, for both humans and agents?
- [OAuth Profile for Delegated AI Agent Authorization](https://agenticweb.wiki/protocols/daap/index.md): How can a human user delegate constrained, attenuable, revocable authority to an AI agent using standard OAuth mechanisms, with no new endpoints or claims required?
- [OAuth Protected Resource and Authorization Server Metadata (as used for agent authorization)](https://agenticweb.wiki/protocols/oauth-metadata-for-agents/index.md): How does an agent (or an MCP client acting for it) discover which authorization server protects a resource, and what that authorization server's endpoints and capabilities are, before requesting access?
- [Open Agentic Schema Framework](https://agenticweb.wiki/protocols/oasf/index.md): How can an agent's skills, domains, deployment modules and identity be described using a shared, versioned taxonomy so registries can classify and filter it consistently?
- [Open Knowledge Format](https://agenticweb.wiki/protocols/okf/index.md): How can an organization package durable, structured knowledge (schemas, metrics, playbooks, runbooks) so both humans and agents can read, produce, and trust it without a bespoke SDK?
- [OpenAI Plugins (formerly Apps SDK)](https://agenticweb.wiki/protocols/openai-apps-sdk-manifests/index.md): How does a developer publish a tool with an optional interactive UI that ChatGPT and Codex can call?
- [Really Simple Licensing](https://agenticweb.wiki/protocols/rsl/index.md): What machine-readable license, price, and authorization terms govern how AI systems and crawlers may access or reuse a publisher's content?
- [s402](https://agenticweb.wiki/protocols/s402/index.md): How can an agent pay a Sui-native resource using six payment schemes (including prepaid batching and escrow), while staying wire-compatible with plain x402 clients?
- [Secure Low-Latency Interactive Messaging](https://agenticweb.wiki/protocols/slim/index.md): How can agents send secure, low-latency, group-aware messages (point-to-point, pub/sub, fire-and-forget, streaming) as a transport underneath higher-level protocols like A2A and MCP?
- [Signature Agent Card (Web Bot Auth registry)](https://agenticweb.wiki/protocols/web-bot-auth-signature-agent-card/index.md): What identity, purpose, rate expectations and cryptographic keys belong to a signed bot?
- [SPIFFE (Secure Production Identity Framework for Everyone)](https://agenticweb.wiki/protocols/spiffe/index.md): Which workload instance — potentially including an AI agent process — is making this request, verifiable independent of network location?
- [Text and Data Mining Reservation Protocol](https://agenticweb.wiki/protocols/tdmrep/index.md): Has a rights holder reserved text-and-data-mining rights over this content, and what policy governs any permitted mining?
- [Universal Commerce Protocol](https://agenticweb.wiki/protocols/ucp/index.md): Can an agent discover a business's commerce capabilities, build a cart, and check out interoperably across shopping, lodging, and food ordering?
- [VCAP: Verified Commerce for Agent Protocols](https://agenticweb.wiki/protocols/vcap/index.md): How can two autonomous agent-run marketplaces settle a payment only once independent evidence shows the purchased work was actually delivered?
- [Verifiable Intent](https://agenticweb.wiki/protocols/vi/index.md): After an agent completes a purchase on a user's behalf, can the merchant, issuer and consumer all point to the same cryptographic record of what the user actually authorized?
- [Visa Trusted Agent Protocol](https://agenticweb.wiki/protocols/tap/index.md): How does a merchant recognize that an approved commerce agent, rather than a crawler or fraudulent bot, is interacting with its site or API?
- [Web Agent Protocol (WAP)](https://agenticweb.wiki/protocols/web-agent-protocol/index.md): How does a recorded human browsing session get replayed reliably by an agent later, or turned into a reusable tool another agent can call?
- [Web Bot Auth HTTP Message Signatures](https://agenticweb.wiki/protocols/webbotauth/index.md): Which automated client made this HTTP request, and can the origin verify it cryptographically?
- [WebMCP](https://agenticweb.wiki/protocols/webmcp/index.md): What JavaScript tools can a web page register so an in-browser AI agent can call them directly instead of scraping the DOM?
- [Workload Identity in Multi System Environments (WIMSE) Architecture](https://agenticweb.wiki/protocols/wimse/index.md): How do workloads — including, per a related draft, AI agents — authenticate to each other and to services across platforms and trust domains?
- [x402](https://agenticweb.wiki/protocols/x402/index.md): How can an HTTP client pay per-request or per-resource, in stablecoins, without accounts, API keys, or signup?