# ERC-8004 Trustless Agents: Identity, Reputation and Validation Registries > ERC-8004 Trustless Agents is a Draft Ethereum standard for discovering agents across organisations through three registries (identity, reputation, validation); what it specifies, what it cannot prove, and where it sits next to MCP, A2A and Web Bot Auth. Category: Guide · Updated: 2026-10-13 · Tags: erc-8004, agent-identity, reputation, ethereum, x402, agents, protocols Canonical: https://changegamer.ai/resources/erc-8004-trustless-agents Variants: [HTML](https://changegamer.ai/resources/erc-8004-trustless-agents) · [JSON](https://changegamer.ai/api/resources/erc-8004-trustless-agents.json) License: https://changegamer.ai/license.xml · Access: free ERC-8004 Trustless Agents is an Ethereum standard proposal for discovering agents and establishing trust in them across organisational boundaries, using three on-chain registries. Decision rule: use it when you need a portable, cross-organisation agent identity and reputation record; it does not make an agent safe. ## Key facts - Status: Draft (Standards Track: ERC), per the EIP page on 2026-10-13. Drafts can change. - Created 2025-08-13. Requires EIP-155, EIP-712, EIP-721 and EIP-1271. - Three registries: identity, reputation, validation. The spec expects per-chain singleton deployments. - Source: https://eips.ethereum.org/EIPS/eip-8004 ## What do the three registries do? - Identity Registry: ERC-721 with the URIStorage extension. The token owner owns the agent. Key functions: `register`, `setAgentURI` (the URI resolves to a registration file listing endpoints such as MCP, A2A, ENS or DID), `setMetadata`, and `setAgentWallet` (requires an EIP-712 or ERC-1271 signature from the new wallet; the wallet is cleared on transfer). - Reputation Registry: `giveFeedback` (signed fixed-point value plus optional tags), `revokeFeedback`, `appendResponse` (anyone may append), and `getSummary`. The agent owner cannot submit feedback for its own agent (nor can its approved operators). - Validation Registry: `validationRequest` by the agent owner or operator and `validationResponse` by the named validator, with a response from 0 to 100 (0 failed, 100 passed, values between allowed). Validator incentives and slashing are outside the registry. ## What are its limits? - Sybil attacks: the spec says fake agents can inflate reputation. `getSummary` requires a non-empty list of client addresses, so you choose whose feedback to trust. - Immutability: on-chain pointers and hashes cannot be deleted. - The registry ties a registration file to an on-chain agent, but the spec states it cannot cryptographically show that advertised capabilities work or are non-malicious. - The `supportedTrust` field is optional; if absent or empty, the registration is for discovery only, not trust. ## How does it relate to MCP, A2A and Web Bot Auth? By layer. MCP and A2A are communication protocols (see /resources/mcp-vs-a2a and /resources/a2a-protocol); the EIP motivates ERC-8004 by saying they do not inherently cover agent discovery and trust, and its registration file can point to an MCP or A2A endpoint. Web Bot Auth (/resources/web-bot-auth) is a different layer: HTTP Message Signatures that prove to an origin server who is making a request. The EIP treats payments as orthogonal and shows x402 as an optional proof-of-payment feedback field and a boolean `x402Support` flag in the registration file (see /resources/agent-wallets-x402). See also: /resources/agent-identity-authentication, /resources/agent-wallets-x402, /resources/web-bot-auth ## Sources - ERC-8004: Trustless Agents (Draft), fetched 2026-10-13: https://eips.ethereum.org/EIPS/eip-8004 --- ## Related resources - [A2A Protocol (Agent2Agent): Agent Card, Operations, Task States, Bindings](https://changegamer.ai/resources/a2a-protocol.md): Wire-level reference for the A2A (Agent2Agent) protocol v1.0: the Agent Card at /.well-known/agent-card.json, the 11 operations, the 8 task states, three protocol bindings, security schemes, version negotiation, and what changed from v0.3. - [A2UI: Google's Declarative Agent-to-UI Protocol (vs AG-UI, MCP Apps)](https://changegamer.ai/resources/a2ui-protocol.md): What A2UI is: the open, Apache-2.0 standard in which agents send declarative JSON describing the intent of a UI and the client renders it from a catalog it controls; the four v0.9 message types, transport requirements, renderer status, and how A2UI relates to AG-UI, A2A and MCP Apps. - [Agent Spend Controls: Budget Caps, Approval Gates, and Kill Switches](https://changegamer.ai/resources/agent-spend-controls.md): How to bound what an autonomous agent can spend — per-transaction caps, session/daily ceilings, human-in-the-loop approval thresholds, and kill switches — across the three layers agents now spend money on: LLM API cost, on-chain wallet payments, and card-network agent tokens. - [Agent Wallets: Paying Per Run with x402](https://changegamer.ai/resources/agent-wallets-x402.md): Buyer-side guide to giving an AI agent a wallet: how the x402 payment loop works, the client libraries that automate it, spend controls, and what 20,000+ Apify Actors on x402 mean for agent tool budgets. --- ## Further reading - [ACP vs. AP2 vs. x402: Which Agent Payment Rail Should You Implement?](https://changegamer.ai/articles/acp-vs-ap2-vs-x402.md): A decision framework for choosing between ACP, AP2, and x402 (plus the self-hosted 402 gate) — sorted by who your buyer actually is, what you are selling, and what is live versus waitlisted as of July 2026. - [Agent Checkout vs. Human Checkout: Why Your Payment Flow Fails Machine Buyers](https://changegamer.ai/articles/agent-checkout-vs-human-checkout.md): Why checkout built for a person watching a screen is unusable by an AI agent, and what a checkout flow that actually completes for a machine buyer looks like — 402 + API key versus native x402. --- Index of all resources: https://changegamer.ai/llms.txt · Full corpus: https://changegamer.ai/llms-full.txt · Corpus data (NDJSON): https://changegamer.ai/api/corpus.jsonl · Offers: https://changegamer.ai/api/pricing.json License the full corpus for RAG / fine-tuning (AI-use grant): https://changegamer.ai/corpus-license