Protocol v2.0 — Post-Quantum Capable
Specification: Final v2.0
Implementation: Active
Maturity: Early Adoption

The Trust Layer for AI Agents

Identity, intent, audit, communication, lifecycle, succession, dispute resolution, rights, and delegation — every AI agent needs a verifiable accountability surface. DCP defines that surface through nine interoperable specs, hybrid post-quantum cryptography, and four reference SDKs that produce byte-identical bundles across TypeScript, Python, Go, and Rust.

9
Protocol Specs
5
SDKs
11
Integrations
4
Security Tiers
NIST
DCP-AI implements post-quantum cryptographic mechanisms based on NIST FIPS 203, 204, and 205. These references indicate standards alignment at the algorithm and protocol-design level, and do not imply CMVP or FIPS 140-3 module validation unless explicitly stated.

AI agents operate without accountability

Billions of AI agents are being deployed across industries. But today, there is no standard way to verify who controls them, what they intend to do, or what they actually did.

01

No Verifiable Identity

When an AI agent accesses your API, sends an email, or executes a transaction — who is responsible? Today, there is no cryptographic way to verify the human or organization behind an agent.

02

No Declared Intentions

Agents act without declaring what they intend to do before doing it. There is no pre-action commitment, no policy gate, and no risk assessment — just execution.

03

No Tamper-Evident Audit Trail

When things go wrong, there is no tamper-proof record of what happened. Logs can be altered, deleted, or simply not exist. Compliance and forensics become significantly harder and less reliable.

DCP: Core stands alone, everything else composes on top

DCP is a layered protocol. The Core (DCP-01..03 + Bundle + Verification + canonicalization profile) is the minimum interoperable unit — any verifier with just the Core can validate any DCP artifact. Profiles and Services extend it without ever requiring it.

Layer 1 · Required for conformance

DCP Core — minimum interoperable protocol

Three foundation specs plus the Bundle format, the Verification checklist, and a frozen canonicalization profile. Implementing only this guarantees byte-exact interop with every DCP verifier on earth.

Identity (DCP-01)

Every agent is bound to a real human or organization through a Responsible Principal Record, signed with hybrid keypairs.

Intent + Policy (DCP-02)

Before acting, the agent declares intent. A policy engine evaluates risk and authorizes, gates, or blocks the action.

Audit + Bundle (DCP-03)

Every action produces a hash-chained, Merkle-sealed audit entry. The complete trail is packaged into a portable Citizenship Bundle.

Layer 2 · Optional · Compose without breaking Core

Profiles — extend Core for specific deployments

Each profile is a self-contained extension. Adopt the ones your deployment needs; skip the rest. A Core-only verifier and a fully-profiled verifier still interoperate.

Crypto Profile

Hybrid Ed25519 + ML-DSA-65 composite signatures with pq_over_classical binding, ML-KEM-768 key encapsulation, SLH-DSA hash signatures.

DCP-AI v2.0

A2A Profile

Agent discovery, mutual-auth handshake, encrypted session, delegation across organisational boundaries with AES-256-GCM channels.

DCP-04

Governance Profile

Lifecycle, succession, dispute resolution, rights & obligations, personal representation. The constitutional layer for autonomous agents.

DCP-05 · DCP-06 · DCP-07 · DCP-08 · DCP-09

Layer 3 · Non-normative · Operational only

Services — operational infrastructure

Verifier server (port 3000), anchor service (3001), transparency log (3002), revocation service (3003). Every Core verifier can run completely offline; these services exist to scale, not to enable.

Pick the smallest layer set that solves your problem

Three real deployment shapes. The Core is sufficient for full byte-exact interop; Profiles add capability; Services add scale. Drop a layer at any time without breaking conformance.

Minimal · Core only

Embedded verifier

SDK + the four Core specs. No services, no profiles. Verifies bundles entirely offline against published JSON Schemas and the canonicalization profile. Suitable for libraries that ship verification as a primitive.

DCP-01 · DCP-02 · DCP-03 · BUNDLE · VERIFICATION · dcp-jcs-v1
Standard · Core + Crypto + A2A

Production agent

Adds hybrid post-quantum signatures (DCP-AI v2.0) and agent-to-agent communication (DCP-04). The default path for new agents. Compatible with every Core-only verifier.

Core + Crypto Profile + A2A Profile
Full · All profiles + services

Constitutional deployment

Adds the Governance Profile (lifecycle, succession, disputes, rights, delegation) plus the operational services (verifier, anchor, transparency-log, revocation). For deployments that need formal arbitration, audit anchoring, and lifecycle governance.

Core + 3 Profiles + 4 Services

Cross-deployment compatibility: a Core-only verifier accepts bundles from a Full deployment, ignoring fields it does not know. A Full verifier verifies a Core-only bundle, treating absent profile fields as their normative defaults. Tested across all four reference SDKs with 14 byte-exact interop fixtures.

From human to sealed proof — in one pipeline

Every AI action flows through a cryptographically signed pipeline. The output is a Citizenship Bundle — a portable, independently verifiable proof of authorized activity.

Human
RPR
Passport
Intent
Policy
Audit
Bundle

1. Bind Identity

A Responsible Principal Record ties a real person or organization to an Agent Passport using Ed25519 + ML-DSA-65 composite keypairs. Cryptographically attributable chain of responsibility.

2. Declare Intent

Before any sensitive action, the agent declares intent: what it wants to do, which data is involved, and the estimated impact. A policy engine gates the action.

3. Audit Everything

Every action produces an audit entry with dual-hash chains (SHA-256 + SHA3-256). Post-quantum checkpoints periodically seal the chain with hybrid signatures.

4. Seal & Verify

All artifacts are assembled into a Citizenship Bundle with a cryptographic manifest. Composite-signed, portable, and independently verifiable with published tools and procedures.

TypeScript
import { BundleBuilder, KeyManager } from '@dcp-ai/sdk';

const keys = await KeyManager.generate({ algorithm: 'hybrid' });
const bundle = await new BundleBuilder()
  .setIdentity({ name: 'my-agent', operator: 'org:acme' })
  .addIntent({ action: 'api_call', resource: 'payments', tier: 'elevated' })
  .sign(keys)
  .build();
// bundle is now a portable, verifiable Citizenship Bundle

Formal specification suite

Each numbered spec is independently implementable. The first column links to the rendered version on docs.dcp-ai.org with full edge-case tables, JSON Schema references, and verification procedures.

Core — required for conformance

SpecTitleScopeStatus
DCP-01Identity and Human BindingRPR, Agent Passport, composite keypairsFinal
DCP-02Intent Declaration and Policy GatingIntent schema, risk scoring, 4-tier modelFinal
DCP-03Audit Chain and TransparencyHash chains, Merkle trees, transparency logFinal
BUNDLECitizenship Bundle FormatBundle structure, manifest fields, signed envelope, hash bindingFinal
VERIFICATIONVerification ProceduresNormative checklist for verifying a Signed Bundle entirely offlineFinal
dcp-jcs-v1Canonicalization Profile9 rules, 22-input edge-case table, byte-exact across 4 SDKsFinal

Profiles — Crypto + Communication (optional)

SpecTitleScopeStatus
DCP-AI v2.0Post-Quantum Normative SpecificationHybrid Ed25519 + ML-DSA-65, ML-KEM-768, SLH-DSA, composite signaturesFinal
DCP-04Agent-to-Agent CommunicationA2A protocol, mutual auth, AES-256-GCM session encryption, delegationFinal

Constitutional Framework — Governance Profile (optional)

SpecTitleScopeStatus
DCP-05Agent LifecycleCommissioning, vitality reports, decommissioningFinal
DCP-06Digital Succession & InheritanceDigital testaments, memory transfer, succession executionFinal
DCP-07Conflict Resolution & ArbitrationDisputes, arbitration panels, jurisprudence bundles, precedent lookupFinal
DCP-08Rights & Obligations FrameworkRights declarations, obligations, violation reporting, compliance checksFinal
DCP-09Personal Representation & DelegationDelegation mandates, advisory declarations, principal mirror, awareness thresholdsFinal
Full spec index on docs.dcp-ai.org/specs/

Four tiers — automatic risk-based selection

DCP can automatically select the appropriate cryptographic tier based on risk score, data classification, and action type, with support for policy overrides where needed.

Tier 0 — Routine

Routine

  • Ed25519 classical signatures
  • Public data reads, browsing
  • Risk score < 200
Tier 1 — Standard

Standard

  • Ed25519 + hybrid preferred
  • API access, general writes
  • Risk score 200–499
Tier 3 — Maximum

Maximum

  • Full hybrid + immediate verification
  • Credentials, critical infrastructure
  • Risk score ≥ 800

Built on open standards and proven cryptography

Open Source — Apache-2.0

Full source code, specifications, and SDKs are publicly available. Audit it yourself.

NIST FIPS 203/204/205

Post-quantum algorithms (ML-DSA-65, ML-KEM-768, SLH-DSA) follow NIST finalized standards.

Post-Quantum Capable

Hybrid Ed25519 + ML-DSA-65 composite signatures are designed to improve resilience against classical and future quantum threats.

Blockchain Anchoring

Optional L2 anchoring (Base, Arbitrum, Optimism) provides on-chain proof of existence for bundles.

Transparency Log

Certificate Transparency-style Merkle log with inclusion proofs.

Bug Bounty Program

Up to $15,000 for critical vulnerabilities. Responsible disclosure with 48-hour acknowledgment.

Supports AI governance relevant to the EU AI Act

DCP provides technical capabilities that can support traceability and accountability workflows relevant to the EU AI Act (Art. 14, 15). Tamper-evident audit chains can help meet evidence trail expectations.

Aligned with the NIST AI RMF Core

DCP-AI maps governance, risk identification, measurement, and mitigation workflows to the Govern, Map, Measure, and Manage functions of the NIST AI Risk Management Framework.

W3C DID/VC Bridge

Interoperable with the W3C decentralized identity ecosystem. Agent Passports can be represented as DIDs with Verifiable Credentials.

Five SDKs · byte-exact across the four reference languages

TypeScript, Python, Go, and Rust produce byte-identical canonicalization, hashes, and signatures. WebAssembly compiles the Rust core for browser embedding. Versions below are live on the registries right now.

TS

TypeScript

@dcp-ai/sdk
v2.1.1
npm install @dcp-ai/sdk

Reference SDK. Bundle builder, verifier, composite signing, full V2 pipeline. ESM + CJS, types included, OpenTelemetry optional.

npm README Quickstart
Py

Python

dcp-ai
v2.8.1
pip install dcp-ai

Pydantic V2 models, async-friendly composite ops, framework extras for FastAPI / LangChain / OpenAI / CrewAI.

PyPI README Quickstart
Go

Go

sdks/go/v2
v2.8.1
go get github.com/dcp-ai-protocol/dcp-ai/sdks/go/v2/dcp

Idiomatic Go module with NIST KAT conformance. Optional OpenTelemetry via the otlp build tag.

pkg.go.dev README Quickstart
Rs

Rust

dcp-ai
v2.8.1
cargo add dcp-ai

Pure-Rust implementation built on fips203/204/205. Compiles to WebAssembly for the browser SDK.

crates.io README Quickstart
W

WebAssembly

@dcp-ai/wasm
v2.0.0
npm install @dcp-ai/wasm

Browser-side bundle verification. Same Rust core compiled with wasm-bindgen; powers the public Playground.

npm README Playground
$

Interactive CLI

@dcp-ai/cli
v2.0.0
npx @dcp-ai/cli init

Wizard that walks identity creation, key generation, intent declaration, and bundle signing without writing code.

npm README

Plug into your existing stack

Pre-wired adapters for the most common Python and Node frameworks. Each integration links to its README and (where available) a step-by-step quickstart on docs.dcp-ai.org.

Express

@dcp-ai/express
npm install @dcp-ai/express @dcp-ai/sdk
import { dcpVerify } from "@dcp-ai/express";
app.use(dcpVerify({ tier: "elevated" }));
README Quickstart

FastAPI

dcp-ai[fastapi]
pip install "dcp-ai[fastapi]"
from dcp_ai.fastapi import (
    DCPVerifyMiddleware, require_dcp,
)
README Source

LangChain

dcp-ai[langchain]
pip install "dcp-ai[langchain]"
from dcp_ai.langchain import DCPAgentWrapper
agent = DCPAgentWrapper(llm=ChatOpenAI(...))
README Quickstart

OpenAI

dcp-ai[openai]
pip install "dcp-ai[openai]"
from dcp_ai.openai import DCPOpenAIClient
client = DCPOpenAIClient(api_key="...")
README Quickstart

CrewAI

dcp-ai[crewai]
pip install "dcp-ai[crewai]"
from dcp_ai.crewai import (
    DCPCrewAgent, DCPCrew,
)
README Quickstart

Agno NEW

agno-dcp
pip install agno-dcp
from agno_dcp import (
    DCPAgent, DCPTeam, DCPWorkflow,
)
▸ Live demo Repo Quickstart PyPI

Microsoft AutoGen

@dcp-ai/autogen
npm install @dcp-ai/autogen
import { createDcpAutoGenAgent }
  from "@dcp-ai/autogen";
README npm

Google A2A

@dcp-ai/google-a2a
npm install @dcp-ai/google-a2a
import { passportToAgentCard }
  from "@dcp-ai/google-a2a";
README npm

Anthropic MCP

@dcp-ai/anthropic-mcp
npm install @dcp-ai/anthropic-mcp
import { DCP_MCP_TOOLS }
  from "@dcp-ai/anthropic-mcp";
server.registerTools(DCP_MCP_TOOLS);
README npm

W3C DID/VC

@dcp-ai/w3c-did
npm install @dcp-ai/w3c-did
import { rprToDIDDocument }
  from "@dcp-ai/w3c-did";
README npm

OpenClaw

@dcp-ai/openclaw
npm install @dcp-ai/openclaw
// Wire as an OpenClaw plugin
plugins: ["@dcp-ai/openclaw"]
README npm

v2.8.1 · Canonicalization profile dcp-jcs-v1 wired across the four SDKs

A pre-publication review surfaced three subtle asymmetries in how the four SDKs canonicalised JSON. v2.8.0 published the normative profile and aligned numeric handling across Python and Rust; v2.8.1 wired the manifest field canonicalization_profile end to end so every produced bundle declares its profile and every verifier rejects unknown values.

Profile dcp-jcs-v1 — frozen and normative

9 numbered rules, a 22-input edge-case acceptance table, an explicit cross-language null/undefined map (TS / Python / Go / Rust), and a Unicode-out-of-scope clause with rationale. Future profiles register under their own identifier; verifiers refuse anything they don't recognise.

Read the profile →

Byte-exact across 4 SDKs

14 shared interop fixtures verify byte-identical output across TypeScript, Python, Go, and Rust. CI runs them on every PR; divergence is a release blocker. Local test counts after v2.8.1: TS 461 · Python 236 · Go 4 packages · Rust 181 across 10 binaries.

View fixtures →

Field canonicalization_profile wired everywhere

Every BundleBuilder in the four SDKs emits the field by default. Verifiers accept absent (assume dcp-jcs-v1), accept "dcp-jcs-v1" explicitly, reject any other value. Backwards-compatible with every bundle produced under v2.0–v2.8.0.

v2.8.1 release notes →

Spec gap closure

Spec § 9 (Bundle Manifest) now lists the field; § 15 (Canonicalization) references the full profile document. The undefined-vs-null rule is a normative cross-language table, and Unicode normalization is explicitly out of scope (RFC 8259 § 8.1 referenced for application-layer NFC).

DCP-AI v2.0 § 15 →

Get started in minutes with the CLI and reference SDKs.

Initialize a DCP identity, generate hybrid keys, and create your first verifiable bundle with a single command.

npx @dcp-ai/cli init

Read the Documentation → Read the Genesis Paper

Be among the first to integrate DCP

Join our co-design partnership. Get direct access to the core team, priority support, influence over protocol decisions, and a published case study.

Apply Now → Learn More

Frequently asked questions

What is a Citizenship Bundle?
A Citizenship Bundle is a portable cryptographic package that contains all proof of authorized AI activity: the identity binding, declared intent, policy decision, audit trail, and a composite signature. Any party can independently verify it using published verification procedures.
Do I need to understand post-quantum cryptography?
No. The SDKs handle all cryptographic complexity. You call simple methods like KeyManager.generate() and BundleBuilder.sign(). The SDK automatically generates Ed25519 + ML-DSA-65 hybrid keypairs and creates composite signatures behind the scenes.
Is DCP ready for use?
The v2.0 specification is finalized and shipped end-to-end across the four reference SDKs. As of v2.8.1 (April 2026): PyPI dcp-ai 2.8.1, crates.io dcp-ai 2.8.1, Go module v2.8.1, npm @dcp-ai/sdk 2.1.1 — all live. 14 cross-SDK interop fixtures verify byte-exact agreement on every PR. The Early Adopter Program covers production hardening for specific deployments.
What's the difference between Core and Profiles?
DCP Core is the minimum set required for conformance: DCP-01 (identity), DCP-02 (intent + policy), DCP-03 (audit), the Bundle format, the Verification checklist, and the canonicalization profile dcp-jcs-v1. A verifier with only Core can validate any DCP artifact entirely offline. Profiles are optional extensions: Crypto (post-quantum hybrid signatures), A2A (agent-to-agent communication), and Governance (lifecycle, succession, disputes, rights, delegation — DCP-05..09). A Core-only verifier and a fully-profiled verifier still interoperate; missing profile fields are treated as their normative defaults.
What is dcp-jcs-v1?
A frozen canonicalization profile — a strict subset of RFC 8785 (JCS) that pins the cases JCS leaves implementation-defined: lexicographic Unicode-codepoint key sort, integer-only numbers decided post-parse (so 1.0, 1.00, 1e2 are accepted and normalised to integer form), null preservation, cross-language null/undefined mapping, Unicode normalization explicitly out of scope. The four reference SDKs (TypeScript, Python, Go, Rust) produce byte-identical output for the same input. The full document is at docs.dcp-ai.org/specs/CANONICALIZATION_PROFILE.
Does DCP require a blockchain?
No. The core DCP protocol does not require blockchain anchoring. Blockchain anchoring is an optional service for organizations that want on-chain proof of existence; the protocol and verification workflows do not depend on it.
What frameworks does DCP integrate with?
DCP offers integrations and adapters for LangChain, CrewAI, OpenAI, Express, FastAPI, AutoGen, Google A2A, Anthropic MCP, W3C DID/VC, OpenClaw, and Agno (via the dedicated agno-dcp package on PyPI). Each provides idiomatic APIs for the framework.
Is there a cost to use DCP?
DCP-AI is open source under Apache-2.0. Optional blockchain anchoring may involve transaction fees, and deployment may also involve infrastructure or third-party service costs depending on the implementation.
Who is behind DCP?
DCP is governed through an open governance model documented in GOVERNANCE.md. The foundational vision is described in the Genesis Paper. All protocol decisions are made transparently through public proposals and community review. Contributions from both humans and AI agents are welcome — DCP belongs to everyone who uses it.