Attestix

Attestation Infrastructure for AI Agents

Verifiable identity, W3C credentials, EU AI Act compliance, delegation chains,
and reputation scoring for every AI agent. 47 MCP tools across 9 modules.

Install

pip install attestix

Why Attestix

On August 2, 2026, the EU AI Act enforcement begins. Fines reach EUR 35M or 7% of global revenue.

Existing compliance tools (Credo AI, Holistic AI, Vanta) are organizational dashboards. None produce machine-readable, cryptographically verifiable proof that an AI agent can present to another agent, regulator, or system.

Agent identity is fragmenting across walled gardens (Microsoft Entra, AWS AgentCore, Google A2A, ERC-8004). No single tool combines agent identity + EU AI Act compliance + verifiable credentials in one protocol.

Attestix fills this gap.

Modules

Quick Start

As an MCP Server (Claude Code)

Add to your Claude Code config (~/.claude.json):

{
  "mcpServers": {
    "attestix": {
      "type": "stdio",
      "command": "python",
      "args": ["/path/to/attestix/main.py"]
    }
  }
}

Then ask Claude:

"Create an identity for my data analysis agent with capabilities: data_analysis, reporting"

As a Python Library

from services.identity_service import IdentityService

svc = IdentityService()
agent = svc.create_identity(
    display_name="MyAgent",
    source_protocol="manual",
    capabilities=["data_analysis", "reporting"],
    description="Analyzes quarterly financial data",
    issuer_name="Acme Corp",
)

print(agent["agent_id"])   # attestix:f9bdb7a94ccb40f1
print(agent["issuer"]["did"])  # did:key:z6Mk...

From Source

git clone https://github.com/VibeTensor/attestix.git
cd attestix
pip install -r requirements.txt
python main.py

EU AI Act Compliance Workflow

Take a high-risk AI agent from zero to fully compliant:

1. create_agent_identity        --> UAIT with DID (Ed25519 signed)
2. record_training_data         --> Article 10 data governance
3. record_model_lineage         --> Article 11 technical documentation
4. create_compliance_profile    --> Risk categorization + obligations
5. record_conformity_assessment --> Article 43 third-party assessment
6. generate_declaration_of_conformity --> Annex V declaration + W3C VC
7. create_verifiable_presentation     --> Signed VP for regulator

High-risk systems are blocked from self-assessment:

record_conformity_assessment(assessment_type="self", ...)
--> ERROR: "High-risk AI systems require third_party conformity assessment"

Full walkthrough: EU AI Act Compliance Guide

How It Works

Every artifact Attestix produces is cryptographically signed with Ed25519:

No cloud dependency. All core operations work offline with local JSON storage.

Architecture

attestix/
  main.py                  # MCP server entry point (47 tools)
  config.py                # Environment-based configuration
  errors.py                # Error handling with JSON logging
  auth/
    signing.py             # Ed25519 key management
    ssrf.py                # SSRF protection for outbound HTTP
  services/
    identity_service.py    # UAIT lifecycle, GDPR erasure
    agent_card_service.py  # A2A agent card operations
    did_service.py         # DID creation and resolution
    delegation_service.py  # UCAN delegation tokens
    reputation_service.py  # Trust scoring
    compliance_service.py  # EU AI Act profiles and assessments
    credential_service.py  # W3C VCs and VPs
    provenance_service.py  # Training data, lineage, audit trail
    blockchain_service.py  # Base L2 anchoring via EAS
  blockchain/
    merkle.py              # Merkle tree for batch anchoring
  tools/                   # MCP tool definitions (one file per module)

All 47 Tools

Standards Conformance

Every standards claim is validated by 91 automated conformance tests that run in Docker alongside the 193 existing tests (284 total). Run them yourself:

docker build -f Dockerfile.test -t attestix-bench . && docker run --rm attestix-bench

Performance (median latency, 1000 runs)

Security

• Ed25519 signatures on all UAITs, VCs, assessments, declarations, and audit entries
• Hash-chained audit trail with SHA-256 for tamper-evident logging
• SSRF protection blocks private IPs, metadata endpoints, and DNS rebinding
• Encrypted key storage with AES-256-GCM when ATTESTIX_KEY_PASSWORD is set
• Private keys never exposed in tool responses
• No external API calls required for core operations

Research Paper

Attestix is described in a research paper covering system architecture, cryptographic pipeline, EU AI Act compliance automation, and evaluation with 284 automated tests.

Attestix: A Unified Attestation Infrastructure for Autonomous AI Agents Pavan Kumar Dubasi, VibeTensor Private Limited, 2026.

Citing Attestix

If you use Attestix in your research, please cite:

@article{dubasi2026attestix,
  title     = {Attestix: A Unified Attestation Infrastructure for Autonomous AI Agents},
  author    = {Dubasi, Pavan Kumar},
  year      = {2026},
  url       = {https://github.com/VibeTensor/attestix},
  note      = {Open-source. Apache License 2.0}
}

Documentation

Full documentation at attestix.vibetensor.com

Disclaimer

Attestix generates machine-readable, cryptographically signed compliance documentation. It is a documentation and evidence tooling system. It does not replace legal counsel, notified body assessments, or official regulatory submissions. Always consult qualified legal professionals for compliance decisions.

Contributing

See CONTRIBUTING.md for development setup and guidelines.

License

Apache License 2.0. See LICENSE.

Built by VibeTensor