OpenA2A AIM (Agent Identity Management) and OpenID Connect
OpenID Connect (OIDC) is an identity layer on top of OAuth 2.0, published by the OpenID Foundation. The current text of its core specification, OpenID Connect Core 1.0 incorporating errata set 2, is dated December 15, 2023. It lets a client, called a Relying Party, verify who signed in and receive claims about them in an ID Token. The person who signs in is the End-User, which Core defines as a human participant. AIM is an open source platform for AI agent identity. It keeps a registry of agents: each agent has an Ed25519 key, a trust score and a list of capabilities checked through an SDK, and MCP servers are attested by the agents that use them.
The two are not alternatives. OpenID Connect signs people in, including the people who deploy and run agents. AIM keeps a registry of the agents. AIM issues its agent tokens through an OAuth 2.0 grant, the JWT-bearer grant type of RFC 7523, and its agent token endpoint returns an access token and no ID Token. In the open source build, the AIM dashboard has no OpenID Connect sign-in.
Checked on against the OpenID specifications and RFCs named in each row and AIM at commit 3ab8285e, opens in a new tab. Each source names the document by its title, the section and the month in the document's header. Core, Discovery and Dynamic Client Registration were read as published with errata set 2, dated December 15, 2023. Where OpenID Connect relies on OAuth 2.0 for a mechanism, the row cites the RFC. To report a line that is wrong, use the contact page.
Capability by capability
Signing people in
OpenID Connect
Defined: signing people in is what OpenID Connect adds to OAuth 2.0. Core says: "OpenID Connect 1.0 is a simple identity layer on top of the OAuth 2.0 protocol. It enables Clients to verify the identity of the End-User based on the authentication performed by an Authorization Server, as well as to obtain basic profile information about the End-User in an interoperable and REST-like manner." The result is an ID Token, "a security token that contains Claims about the Authentication of an End-User by an Authorization Server when using a Client, and potentially other requested Claims."
Source: OpenID Connect Core 1.0 incorporating errata set 2, December 2023, opens in a new tab, OpenID Connect Core 1.0 incorporating errata set 2, section 2, December 2023, opens in a new tab
AIM
The dashboard has local email and password sign-in. No single sign-on, OpenID Connect or SAML login is wired in the open source build: the repository contains Google, Microsoft and Okta provider code, but no login route calls it.
Who an ID Token identifies
OpenID Connect
An End-User, which Core defines as a "Human participant." The ID Token's sub claim is "A locally unique and never reassigned identifier within the Issuer for the End-User", and its aud claim "MUST contain the OAuth 2.0 client_id of the Relying Party as an audience value." The client that receives it is a Relying Party, an "OAuth 2.0 Client application requiring End-User Authentication and Claims from an OpenID Provider." Dynamic Client Registration 1.0 defines how a Relying Party registers and obtains "the OAuth 2.0 Client ID for this Relying Party." For a client acting on its own behalf, with no End-User, OAuth 2.0 defines the client credentials grant. None of the specifications cited on this page uses the term AI agent.
Source: OpenID Connect Core 1.0 incorporating errata set 2, section 1.2, December 2023, opens in a new tab, OpenID Connect Core 1.0 incorporating errata set 2, section 2, December 2023, opens in a new tab, OpenID Connect Dynamic Client Registration 1.0 incorporating errata set 2, December 2023, opens in a new tab, RFC 6749, section 4.4, October 2012, opens in a new tab
AIM
Each agent is registered with its own Ed25519 public key. The token endpoint accepts one grant: a JWT-bearer assertion signed with that key, in exchange for a bearer token for AIM's own agent API. AIM does not implement the client credentials grant: the token endpoint refuses it.
Source: agent_service.go, opens in a new tab, oauth_token_handler.go, opens in a new tab
Client authentication with keys
OpenID Connect
Defined. Core lists the methods a client can use to authenticate at the token endpoint: client_secret_basic, client_secret_post, client_secret_jwt, private_key_jwt and none. Under private_key_jwt, "Clients that have registered a public key sign a JWT using that key." Dynamic Client Registration 1.0 lets a client publish its keys at a jwks_uri: "If the Client signs requests to the Server, it contains the signing key(s) the Server uses to validate signatures from the Client."
Source: OpenID Connect Core 1.0 incorporating errata set 2, section 9, December 2023, opens in a new tab, OpenID Connect Dynamic Client Registration 1.0 incorporating errata set 2, section 2, December 2023, opens in a new tab
AIM
An agent can sign each request with its Ed25519 key. On routes behind the request-signature check, the server verifies the signature over the method, path, timestamp and body, and rejects a timestamp more than 30 seconds from its own clock. The action-check request is signed over the action's fields (agent, capability, resource, context and timestamp) and verified against the agent's registered key. Agent routes also accept an API key or a bearer token in place of a signature.
Authentication flows
OpenID Connect
Core defines three: "the Authorization Code Flow (response_type=code), the Implicit Flow (response_type=id_token token or response_type=id_token), or the Hybrid Flow". Each starts at the authorization endpoint, and one step of the Authorization Code Flow reads: "Authorization Server sends the End-User back to the Client with an Authorization Code." The Client-Initiated Backchannel Authentication Flow (CIBA), a separate final specification, has "direct Relying Party to OpenID Provider communication without redirects through the user's browser". In CIBA, "The user starts the flow with the Relying Party at the Consumption Device, but authenticates and grants consent on the Authentication Device."
Source: OpenID Connect Core 1.0 incorporating errata set 2, section 3, December 2023, opens in a new tab, OpenID Connect Core 1.0 incorporating errata set 2, section 3.1.1, December 2023, opens in a new tab, OpenID Connect Client-Initiated Backchannel Authentication Flow: Core 1.0, September 2021, opens in a new tab
AIM
The agent token endpoint accepts one grant type, urn:ietf:params:oauth:grant-type:jwt-bearer, and refuses any other with unsupported_grant_type. It reads the JWT from client_assertion, the parameter RFC 7523 defines for client authentication, and not from assertion, the parameter RFC 7523 defines for this grant. The JWT must be signed with the agent's registered Ed25519 key, its sub must equal client_id, and it must carry an exp claim and an aud claim that names the AIM server. The response is a bearer token with expires_in set to 3600.
Source: oauth_token_handler.go, opens in a new tab, validateAssertionClaims, opens in a new tab
Token exchange and delegation
OpenID Connect
Not found in the OpenID specifications cited on this page, checked 2026-10-16: none defines a client passing its authority on to another client. Nearest defined mechanisms: CIBA, which "allows a Relying Party that has an identifier for a user to obtain tokens from the OpenID Provider" once the user approves on another device, and OAuth 2.0 token exchange, whose RFC 8693 lists among the token types a request can carry urn:ietf:params:oauth:token-type:id_token, which "Indicates that the token is an ID Token as defined in Section 2 of [OpenID.Core]."
Source: OpenID Connect Client-Initiated Backchannel Authentication Flow: Core 1.0, September 2021, opens in a new tab, RFC 8693, section 3, January 2020, opens in a new tab
AIM
OAuth token exchange (RFC 8693) is not implemented: the agent token endpoint accepts the JWT-bearer grant only. Delegation exists outside OAuth, in the TypeScript SDK, which creates and verifies Ed25519-signed delegation chains between keys. The AIM server does not issue or check them.
Source: oauth_token_handler.go, opens in a new tab, delegation.ts, opens in a new tab
Proof of possession
OpenID Connect
Not defined in Core. At the UserInfo endpoint, "The Access Token obtained from an OpenID Connect Authentication Request MUST be sent as a Bearer Token". CIBA says "Any sender constraining method that relies on the Client presenting key material (i.e a certificate or a public key) at the Token Endpoint can be adapted to work with CIBA." The mechanisms themselves are OAuth 2.0 extensions: DPoP in RFC 9449, and access tokens bound to a mutual-TLS client certificate in RFC 8705.
Source: OpenID Connect Core 1.0 incorporating errata set 2, section 5.3.1, December 2023, opens in a new tab, OpenID Connect Client-Initiated Backchannel Authentication Flow: Core 1.0, section 5, September 2021, opens in a new tab, RFC 9449, September 2023, opens in a new tab, RFC 8705, February 2020, opens in a new tab
AIM
DPoP is not found in the AIM repository at this commit. Tokens from the token endpoint are plain bearer tokens. Requests authenticated by signature are signed per call, as in the row on client authentication.
Permissions
OpenID Connect
Defined as OAuth 2.0 scope, which Core uses for access tokens: "The scopes associated with Access Tokens determine what resources will be available when they are used to access OAuth 2.0 protected endpoints." Core adds scope values that request claims about the End-User: profile, which "requests access to the End-User's default profile Claims", and email, address and phone. Claims beyond the standard set are allowed: "other Claims MAY be used in conjunction with the standard Claims."
Source: OpenID Connect Core 1.0 incorporating errata set 2, section 5.4, December 2023, opens in a new tab, OpenID Connect Core 1.0 incorporating errata set 2, section 5.1.2, December 2023, opens in a new tab
AIM
Each agent has a list of granted capabilities. The SDK, running inside the agent process, sends each wrapped action to the server, which compares the capability name with that list. In the default mode (monitoring) a capability outside the list is allowed and, under the default policy, raises an alert. The Python SDK's decorator also registers a capability on first use, and in this mode the server grants it automatically. In strict mode a capability outside the list is denied under the default policy, and the SDK blocks the call.
Source: agent_service.go, opens in a new tab, enforcement.py, opens in a new tab
Registry of agents with a trust score
OpenID Connect
Not found in the OpenID specifications cited on this page, checked 2026-10-16: none defines a risk or trust score for a client. Nearest defined mechanisms: client registration, in the row on who an ID Token identifies, and Trust Marks in OpenID Federation 1.0, "statements of conformance to sets of criteria determined by an accreditation authority". The specification says: "Trust Marks used by this specification are signed JWTs."
Source: OpenID Connect Dynamic Client Registration 1.0 incorporating errata set 2, December 2023, opens in a new tab, OpenID Federation 1.0, section 7, February 2026, opens in a new tab
AIM
Each agent has a trust score calculated from nine weighted factors; a factor with no data is left out of the weighted sum. It is recalculated on events such as registration, verification by an administrator, capability changes and capability reports from the SDK. A capability violation also lowers the stored score directly. Each action check reads the stored score.
MCP servers
OpenID Connect
MCP is not mentioned in the OpenID specifications cited on this page. The MCP specification, revision 2026-07-28, lists OpenID Connect Discovery 1.0 and OpenID Connect Dynamic Client Registration 1.0 among the specifications its authorization is based on. "MCP authorization servers MUST provide at least one of the following discovery mechanisms": OAuth 2.0 Authorization Server Metadata (RFC 8414) or OpenID Connect Discovery 1.0. And "MCP clients MUST support both discovery mechanisms to obtain the information required to interact with the authorization server." Authorization is optional for an MCP implementation.
Source: MCP specification 2026-07-28, Authorization, opens in a new tab, OpenID Connect Discovery 1.0 incorporating errata set 2, December 2023, opens in a new tab
AIM
AIM is not an OAuth authorization server for MCP. Agents submit signed attestations of the MCP servers they connect to, and the server checks each signature against the agent's registered key. A drift alert is raised when a capability detection run finds that a server's manifest has changed since the last run.
Source: mcp_attestation_service.go, opens in a new tab, mcp_manifest_service.go, opens in a new tab
Workload identity
OpenID Connect
Not found in the OpenID specifications cited on this page, checked 2026-10-16: none names SPIFFE or a Kubernetes service account token. Nearest defined mechanism: Discovery 1.0, under which an issuer publishes its metadata, including a jwks_uri whose document "contains the signing key(s) the RP uses to validate signatures from the OP."
AIM
Not implemented for agents at this commit: AIM accepts no SPIFFE identity, Kubernetes service account token or external OpenID Connect issuer as an agent credential.
Source: oauth_token_handler.go, opens in a new tab, pqc_agent_auth.go, opens in a new tab
Audit
OpenID Connect
Not found in the OpenID specifications cited on this page, checked 2026-10-16: none defines an audit log. Nearest defined mechanism: the ID Token, which records who signed in, at which issuer, for which client and when, with auth_time, "Time when the End-User authentication occurred", and amr, "identifiers for authentication methods used in the authentication."
Source: OpenID Connect Core 1.0 incorporating errata set 2, section 2, December 2023, opens in a new tab
AIM
An action check that passes its signature check is written to the audit log as a row with the agent, the capability, the resource, the stored trust score and whether it was allowed. The write is best effort: if it fails, the check still returns its decision. The log is not append-only at the database layer and is not signed.
Source: verification_handler.go, opens in a new tab, audit_log_repository.go, opens in a new tab, SECURITY.md, opens in a new tab
License and hosting
OpenID Connect
A set of specifications, not software: there is no server to install or host. Core, Discovery and Dynamic Client Registration are final specifications of the OpenID Foundation, read here as published with errata set 2 in December 2023. CIBA Core 1.0 is a final specification of September 2021. The OpenID Providers and Relying Party libraries that implement them are separate software.
Source: OpenID Connect Core 1.0 incorporating errata set 2, December 2023, opens in a new tab, OpenID Connect Discovery 1.0 incorporating errata set 2, December 2023, opens in a new tab, OpenID Connect Dynamic Client Registration 1.0 incorporating errata set 2, December 2023, opens in a new tab, OpenID Connect Client-Initiated Backchannel Authentication Flow: Core 1.0, September 2021, opens in a new tab
AIM
The repository is Apache-2.0. Self-hosted with Docker Compose. Also offered as a hosted service, AIM Cloud, which is built from a private mirror of the repository with cloud-only additions.
Source: LICENSE, opens in a new tab, docs/DEPLOYMENT.md, opens in a new tab, docker-compose.quickstart.yml, opens in a new tab
Which layer each one covers
OpenID Connect
- Signing people in to applications and returning an ID Token about each sign-in.
- Publishing an OpenID Provider's endpoints and signing keys.
- Registering clients and the keys they authenticate with.
- Signing a person in on a second device when the client has no browser session with them.
AIM
- Keeping a registry of agents, each with its own key.
- Scoring each agent and recording the action checks it submits.
- Checking each action an agent submits against its granted capabilities.
- Recording which MCP servers agents attest to, and alerting when a detection run finds a server's manifest changed.
A team can run both: an OpenID Provider for the people who sign in to its applications and tools, AIM for its registry of agents and their action checks. AIM has no OpenID Connect integration at this commit: it accepts no token from an external OpenID Connect issuer as an agent credential, and no OpenID Connect sign-in is wired in to its dashboard in the open source build. Agents authenticate to AIM with their own keys or an AIM API key, and in the open source build people sign in to the AIM dashboard with a local account.
Registering an agent with AIM
In the Python SDK, secure() registers an agent and perform_action wraps the actions to be checked. secure() needs credentials first, from aim-sdk login or an SDK downloaded from the dashboard. For a self-hosted server, AIM's README at this commit says to create the agent in the dashboard first. The capability name on the decorator is what the server compares with the agent's grants. The resource is recorded with the check and is not compared.
from aim_sdk import secure
agent = secure("my-agent")
@agent.perform_action("db:read", resource="users_table")
def get_user_data(user_id):
...