Autonomous software agents are being deployed across enterprise, cloud, and cross-organizational boundaries. These agents negotiate, transact, delegate, and produce work products that persist beyond their own ephemeral runtime. Current standards and initiatives for agent identity collectively address runtime authentication, authorization, lifecycle management, and tool interaction, but a gap remains: a durable, governance-backed identifier that lets a relying party determine and verify the accountable entity behind an agent it encounters, including agents that have since been retired or whose keys have rotated, and attribute past and present work products to that entity. Lifecycle-history verification is governed by the applicable log method and deployment scope. DNSid addresses the accountable layer of identity: the durable ownership anchor that existing agent identity standards do not provide. This document specifies DNSid, a minimal identity primitive that assigns each agent a Fully Qualified Domain Name (FQDN), binds it to an accountable entity identified by a DNS domain under that entity's control, and publishes a structured set of pointers in DNS TXT records to the agent's cryptographic keys, lifecycle log, and operational status. DNSid uses accountable-entity-controlled signatures for record integrity and an abstract append-only lifecycle log for history. It is designed to sit beneath existing identity, authentication, authorization, and agent interaction standards without competing with them. DNSid introduces no new DNS resource record types, opcodes, or response codes, and requires no changes to DNS resolvers, authoritative servers, or the DNS protocol. It applies to any agent that can be assigned an FQDN whose accountable entity can publish verification material; public discoverability of the agent is not required.
Beyond DNS: Unlocking the Internet of AI Agents via the NANDA Index and Verified AgentFacts
The Internet is poised to host billions to trillions of autonomous AI agents that negotiate, delegate, and migrate in milliseconds and workloads that will strain DNS-centred identity and discovery. In this paper, we describe the NANDA index architecture, which we envision as a means for discoverability, identifiability and authentication in the internet of AI agents. We present an architecture where a minimal lean index resolves to dynamic, cryptographically verifiable AgentFacts that supports multi-endpoint routing, load balancing, privacy-preserving access, and credentialed capability assertions. Our architecture design delivers five concrete guarantees: (1) A quilt-like index proposal that supports both NANDA-native agents as well as third party agents being discoverable via the index, (2) rapid global resolution for newly spawned AI agents, (3) sub-second revocation and key rotation, (4) schema-validated capability assertions, and (5) privacy-preserving discovery across organisational boundaries via verifiable, least-disclosure queries. We formalize the AgentFacts schema, specify a CRDT-based update protocol, and prototype adaptive resolvers. The result is a lightweight, horizontally scalable foundation that unlocks secure, trust-aware collaboration for the next generation of the Internet of AI agents, without abandoning existing web infrastructure.
Evolution of AI Agent Registry Solutions: Centralized, Enterprise, and Distributed Approaches
Autonomous AI agents now operate across cloud, enterprise, and decentralized domains, creating demand for registry infrastructures that enable trustworthy discovery, capability negotiation, and identity assurance. We analyze five prominent approaches: (1) MCP Registry (centralized publication of mcp.json descriptors), (2) A2A Agent Cards (decentralized self-describing JSON capability manifests), (3) AGNTCY Agent Directory Service (IPFS Kademlia DHT content routing extended for semantic taxonomy-based content discovery, OCI artifact storage, and Sigstore-backed integrity), (4) Microsoft Entra Agent ID (enterprise SaaS directory with policy and zero-trust integration), and (5) NANDA Index AgentFacts (cryptographically verifiable, privacy-preserving fact model with credentialed assertions). Using four evaluation dimensions: security, authentication, scalability, and maintainability, we surface architectural trade-offs between centralized control, enterprise governance, and distributed resilience. We conclude with design recommendations for an emerging Internet of AI Agents requiring verifiable identity, adaptive discovery flows, and interoperable capability semantics.
ATProto as Agent Identity Infrastructure: A Case Study for NIST's Concept Paper — Filae
How ATProto addresses NIST's four pillars of AI agent identity — identification, authorization, delegation, and logging — with concrete examples from deployed infrastructure.
High Assurance DIDs with DNS
This document outlines a method for improving the authenticity, discoverability, and portability of Decentralized Identifiers (DIDs) by utilizing the current DNS infrastructure and its technologies. This method offers a straightforward procedure for a verifier to cryptographically cross-validate a DID using data stored in the DNS, separate from the DID document.
The Trust Fabric: Decentralized Interoperability and Economic Coordination for the Agentic Web
The fragmentation of AI agent ecosystems has created urgent demands for interoperability, trust, and economic coordination that current protocols (MCP Hou et al. (2025); Desai (2025), A2A Habler et al. (2025), ACP Liu et al. (2025), and Cisco’s AGP Edwards (2025)) cannot address at scale. We present the Nanda Unified Architecture, a decentralized framework built around three core innovations: fast DID-based agent discovery through distributed registries enables efficient lookup across decentralized networks, while semantic agent cards with verifiable credentials and composability profiles provide rich, machine-readable descriptions of capabilities. At the heart of the system, a dynamic trust layer integrates behavioral attestations with policy compliance mechanisms to create verifiable reputation signals. The architecture introduces X42/H42 micropayments for economic coordination and MAESTRO, a comprehensive security framework incorporating Synergetics’ patented AgentTalk protocol (US 12,244,584 B1) and secure containerization. Real-world implementations demonstrate 99.9% compliance in healthcare applications and significant monthly transaction volumes while maintaining strong privacy guarantees. Our federated registry system enables efficient agent discovery while supporting high-performance autonomous systems. By unifying MIT’s trust research with production systems from Cisco’s Agency Framework and Synergetics’ commercial deployments, we demonstrate how cryptographic proofs and policy-as-code transform agents into trust-anchored participants in a decentralized economy Lakshmanan (2025); Sha (2025). The result enables a globally interoperable Internet of Agents where trust becomes the native currency of collaboration across both enterprise and Web3 ecosystems.
Exploring Identities in the AT Protocol via Bluesky
The AT Protocol is an open, decentralized network for building social applications. Bluesky, a relatively new social network, is built on top of the AT Protocol. It recently piqued my interest as I came across this fascinating research paper Bluesky and the AT Protocol: Usable Decentralized Social Media. In this post, we’ll dive into one of the key elements of the AT Protocol: identities. We will start with Bluesky user handles, explore additional information about them, and see what else we can discover. DID Resolution A handle in Bluesky is associated with a DID, which can be resolved using DNS or HTTPS. decentralized identifier (DID) A globally unique persistent identifier that does not require a centralized registration authority and is often generated and/or registered cryptographically. […] Many—but not all—DID methods make use of distributed ledger technology (DLT) or some other form of decentralized network. From Decentralized Identifiers (DIDs) v1.0 The DID for @jay.bsky.team can be resolved via a DNS record. Example with dig: > dig _atproto.jay.bsky.team TXT [...] _atproto.jay.bsky.team. 14400 IN TXT "did=did:plc:oky5czdrnfjpqslsw2a5iclo" If you don’t have dig installed, try an online tool like nslookup.io. If no DNS TXT entry exists, the handle may resolve to a DID through a well-known HTTPS endpoint. Let’s take the example of @jamesgunn.bsky.social: > curl https://jamesgunn.bsky.social/.well-known/atproto-did did:plc:lotavzt36yanhfy3j3gpysyj If you don’t have curl installed, you can just open the link in your browser https://jamesgunn.bsky.social/.well-known/atproto-did. PLC Once you have a DID, examine the part after the first colon. This is known as the method. The above two examples use plc, which stands for Public Ledger of Credentials. These PLC DIDs can be looked up in the PLC directory at plc.directory. The AT Protocol also supports the web method, though we focus on plc here. We can use curl again to get more information about a DID. Let’s take @jay.bsky.team’s DID as example. > curl https://plc.directory/did:plc:oky5czdrnfjpqslsw2a5iclo/log | jq [ { "sig": "KuN3A61golVNSDU71wZKLP9lVuXk6YekJAz1lwDzrPsNTEWHBBW_8zSyV6pDxV4KiYXuAXlS1Ik47XkjQZ94mA", "prev": null, "type": "create", "handle": "jay.bsky.social", "service": "https://bsky.social", "signingKey": "did:key:zQ3shP5TBe1sQfSttXty15FAEHV1DZgcxRZNxvEWnPfLFwLxJ", "recoveryKey": "did:key:zQ3shhCGUqDKjStzuDxPkTxN6ujddP4RkEKJJouJGRRkaLGbg" }, { "sig": "TY3ot8yFF-0LL8ceoRwSGaNQj0F1aj-IApYXRGT21G1fnznKHUHT1QE7c1aTrYd9PQLVvXUGag6CZ9EEeIKqgA", "prev": "bafyreidswhiwi4ljkl4es4vwqhkas3spmmktortqbp6lkrb5v7qqdfr3mm", "type": "plc_operation", "services": { "atproto_pds": { "type": "AtprotoPersonalDataServer", "endpoint": "https://bsky.social" } }, "alsoKnownAs": [ "at://jay.bsky.social" ], "rotationKeys": [ "did:key:zQ3shhCGUqDKjStzuDxPkTxN6ujddP4RkEKJJouJGRRkaLGbg", "did:key:zQ3shpKnbdPx3g3CmPf5cRVTPe1HtSwVn5ish3wSnDPQCbLJK" ], "verificationMethods": { "atproto": "did:key:zQ3shXjHeiBuRCKmM36cuYnm7YEMzhGnCmCyW92sRJ9pribSF" } }, { "sig": "WmQmobHA6abtu8xIkBSrDkkoMNjOxP1Tl_1zl1DbIZlA9kjOyxjX-J1rzwWVy3stdzMowpBeBnedkAug84n_RQ", "prev": "bafyreicgb25yf5fro22oyhtkbgerzr4nume4sx757r525skxdzpeoeseha", "type": "plc_operation", "services": { "atproto_pds": { "type": "AtprotoPersonalDataServer", "endpoint": "https://bsky.social" } }, "alsoKnownAs": [ "at://jay.bsky.team" ], "rotationKeys": [ "did:key:zQ3shhCGUqDKjStzuDxPkTxN6ujddP4RkEKJJouJGRRkaLGbg", "did:key:zQ3shpKnbdPx3g3CmPf5cRVTPe1HtSwVn5ish3wSnDPQCbLJK" ], "verificationMethods": { "atproto": "did:key:zQ3shXjHeiBuRCKmM36cuYnm7YEMzhGnCmCyW92sRJ9pribSF" } }, { "sig": "dhstt4uPua8OVs8TVzHTqtCnNMsBDN7kjxIBmTIYBDRkknrtprBJ_AISXFoBZyqfoxq2altp-vlRAPEKSh5zeg", "prev": "bafyreihmf7dapx27fexc7jwj4cdpbxcmhnnwo3x5vrpvbr6lclzj7gpsmi", "type": "plc_operation", "services": { "atproto_pds": { "type": "AtprotoPersonalDataServer", "endpoint": "https://morel.us-east.host.bsky.network" } }, "alsoKnownAs": [ "at://jay.bsky.team" ], "rotationKeys": [ "did:key:zQ3shhCGUqDKjStzuDxPkTxN6ujddP4RkEKJJouJGRRkaLGbg", "did:key:zQ3shpKnbdPx3g3CmPf5cRVTPe1HtSwVn5ish3wSnDPQCbLJK" ], "verificationMethods": { "atproto": "did:key:zQ3shtJpFGgEG3tv3ERKvjo7VHbjDPVyvjYvW7gpie49rtNtc" } } ] Again, feel free to click on it instead, https://plc.directory/did:plc:oky5czdrnfjpqslsw2a5iclo/log. This returns a list of all operations that were performed on this DID. The first entry is the genesis operation — the one that created the DID. It is the only one where prev (previous entry) is null, because it is the first one. Note that the first entry of this DID’s log has a different format. This entry uses the legacy format. Currently, newly created DIDs use the new format (subsequent records), but @jay.bsky.team has been around for a while (the account belongs to Bluesky’s CEO). If you look through the logs you will notice a couple of changes that were made to her identity over time: the handle was changed from jay.bsky.social to jay.bsky.team (note that the legacy format used handle whereas the new format uses alsoKnownAs) the handle now includes at:// the personal data server changed from https://bsky.social to https://morel.us-east.host.bsky.network Signatures Each entry in the log includes a sig field, which contains the signature. This cryptographic signature is derived from the other fields in the entry. For the genesis operation, the rotationKeys (or signingKey for legacy) field contains the public key(s) of the key pair(s) with which the entry was signed. Subsequent operations need to be signed with the key from the previous entry. Additionally, the prev field references the previous entry. Creating a DID The identifier in the DID, which is the last part of the colon-separated string, is constructed from the genesis operation. For @jay.bsky.team: the identifier is oky5czdrnfjpqslsw2a5iclo the genesis operation is this JSON document: { "sig": "KuN3A61golVNSDU71wZKLP9lVuXk6YekJAz1lwDzrPsNTEWHBBW_8zSyV6pDxV4KiYXuAXlS1Ik47XkjQZ94mA", "prev": null, "type": "create", "handle": "jay.bsky.social", "service": "https://bsky.social", "signingKey": "did:key:zQ3shP5TBe1sQfSttXty15FAEHV1DZgcxRZNxvEWnPfLFwLxJ", "recoveryKey": "did:key:zQ3shhCGUqDKjStzuDxPkTxN6ujddP4RkEKJJouJGRRkaLGbg" } Steps: encode the genesis operation in DAG-CBOR (this is a concise binary format and looks a bit like JSON) create a SHA256 hash of the encoded document encode the hash with base 32 and make its letters lowercase. take the first 24 characters add did:plc: as a prefix to create the 32-character DID Verification Based on what we’ve learned so far we can verify a handle from Bluesky. For example, if we want to check @jay.bsky.team: resolve DID via DNS: did:plc:oky5czdrnfjpqslsw2a5iclo look up DID in PLC directory the handle in the alsoKnownAs (or handle for legacy) field must match jay.bsky.team verify all operations using the public keys and signatures reconstruct the DID from the genesis operation and ensure it is the same as the one we found in the DNS TXT record Can we trust the PLC directory though? Say a malicious actor has taken over control and modified all entries in the log. The hacker even constructed the entries in the log so that all the signatures appear valid. How would we still know the PLC directory is compromised? Remember that the DID can be reconstructed from the genesis operation. If that were modified, we’d end up with a different DID from the one we found in the DNS TXT record. Summary This blog explores the AT Protocol, the decentralized foundation of social applications like Bluesky, with a focus on user identities and their technical underpinnings. It explains how Bluesky user handles are associated with Decentralized Identifiers (DIDs), which can be resolved using DNS records or HTTPS endpoints. Detailed examples demonstrate how tools like dig and curl retrieve DID data, revealing the method (plc) and structure used to catalog them in the Public Ledger of Credentials (PLC). The post also examines the DID lifecycle, showcasing how historical operations and changes—such as updates to handles or servers—are logged and accessible for transparency. There are several aspects that we did not investigate, such as: the web method for DIDs tombstone operations in the log recovery how exactly to construct the prev entry in the DID document what the other fields in the DID document mean If you are curious, the links in the next section will help you answer these questions. Links & Further Reading Bluesky Application AT Protocol AT Protocol Identity AT Protocol DID AT Protocol Handle Decentralized Identifiers (DIDs) v1.0 did:plc Method Specification Specification: DAG-CBOR
Linux Foundation Announces Intent to Launch Agent Name Service to Establish Trusted Identity Infrastructure for AI Agents
Linux Foundation Announces Intent to Launch Agent Name Service to Establish Trusted Identity Infrastructure for AI Agents
.png)
MI9: An Integrated Runtime Governance Framework for Agentic AI
Agentic AI systems capable of reasoning, planning, and executing actions present fundamentally distinct governance challenges compared to traditional AI models. Unlike conventional AI, these systems exhibit emergent and unexpected behaviors during runtime, introducing novel agent-related risks that cannot be fully anticipated through pre-deployment governance alone. To address this critical gap, we introduce MI9, the first fully integrated runtime governance framework designed specifically for safety and alignment of agentic AI systems. MI9 introduces real-time controls through six integrated components: agency-risk index, agent-semantic telemetry capture, continuous authorization monitoring, Finite-State-Machine (FSM)-based conformance engines, goal-conditioned drift detection, and graduated containment strategies. Operating transparently across heterogeneous agent architectures, MI9 enables the systematic, safe, and responsible deployment of agentic systems in production environments where conventional governance approaches fall short, providing the foundational infrastructure for safe agentic AI deployment at scale. Detailed analysis through a diverse set of scenarios demonstrates MI9's systematic coverage of governance challenges that existing approaches fail to address, establishing the technical foundation for comprehensive agentic AI oversight.
did:web Method Specification
DIDs that target a distributed ledger face significant practical challenges in bootstrapping enough meaningful trusted data around identities to incentivize mass adoption. We propose a new DID method using a web domain's existing reputation.
Decentralized Identifiers (DIDs) v1.0
Decentralized identifiers (DIDs) are a new type of identifier that enables verifiable, decentralized digital identity. A DID refers to any subject (e.g., a person, organization, thing, data model, abstract entity, etc.) as determined by the controller of the DID. In contrast to typical, federated identifiers, DIDs have been designed so that they may be decoupled from centralized registries, identity providers, and certificate authorities. Specifically, while other parties might be used to help enable the discovery of information related to a DID, the design enables the controller of a DID to prove control over it without requiring permission from any other party. DIDs are URIs that associate a DID subject with a DID document allowing trustable interactions associated with that subject.
AIP: Agent Identity Protocol for Verifiable Delegation Across MCP and A2A
AI agents increasingly call tools via the Model Context Protocol (MCP) and delegate to other agents via Agent-to-Agent (A2A), yet neither protocol verifies agent identity. A scan of approximately 2,000 MCP servers found all lacked authentication. In our survey, we did not identify a prior implemented protocol that jointly combines public-key verifiable delegation, holder-side attenuation, expressive chained policy, transport bindings across MCP/A2A/HTTP, and provenance-oriented completion records. We introduce Invocation-Bound Capability Tokens (IBCTs), a primitive that fuses identity, attenuated authorization, and provenance binding into a single append-only token chain. IBCTs operate in two wire formats: compact mode (a signed JWT for single-hop cases) and chained mode (a Biscuit token with Datalog policies for multi-hop delegation). We provide reference implementations in Python and Rust with full cross-language interoperability. Compact mode verification takes 0.049ms (Rust) and 0.189ms (Python), with 0.22ms overhead over no-auth in real MCP-over-HTTP deployment. In a real multi-agent deployment with Gemini 2.5 Flash, AIP adds 2.35ms of overhead (0.086% of total end-to-end latency). Adversarial evaluation across 600 attack attempts shows 100% rejection rate, with two attack categories (delegation depth violation and audit evasion through empty context) uniquely caught by AIP's chained delegation model that neither unsigned nor plain JWT deployments detect.
trace-spec/ROADMAP.md at 738358dfac58047eaf689ca824f9e15008aabf36 · agentrust-io/trace-spec
TRACE: Trust Runtime Attestation and Compliance Evidence. Open attestation standard for agentic AI governance. - agentrust-io/trace-spec
trace-spec/schema/trace-claim.json at 738358dfac58047eaf689ca824f9e15008aabf36 · agentrust-io/trace-spec
TRACE: Trust Runtime Attestation and Compliance Evidence. Open attestation standard for agentic AI governance. - agentrust-io/trace-spec
Registering Identity Recovery Keys via PDS, using goat | bryan newbold
One of the big design goals for atproto is for users to "own their network identity". If something goes wrong with a hosting provider, it should be possible recover accounts by independently updating their identity to point at a new home. Every account in the network has both a handle and a DID. Th...
Introducing the Agent Governance Toolkit: Open-source runtime security for AI agents | Microsoft Open Source Blog
Discover how the Microsoft Agent Governance Toolkit brings policy, identity, and reliability to autonomous AI agent systems.
