asqav

What is asqav?

asqav is an open-source Python SDK that brings governance and compliance to AI agent workflows. It attaches quantum-safe cryptographic signatures to every agent action, creating tamper-proof audit trails that hold up under regulatory scrutiny. If you build AI agents with LangChain, CrewAI, or OpenAI, asqav helps you track what your agents do and enforce rules around their behavior.

Top Features:

• Quantum-Safe Signatures: every agent action is signed using the ML-DSA-65 FIPS 204 standard.
• Hash-Chained Audit Trails: actions link into a tamper-proof chain for full traceability.
• Policy Enforcement: define rules that block or allow specific agent actions in real time.

Use Cases:

• Compliance Reporting: generate audit reports that satisfy EU AI Act requirements automatically.
• Multi-Party Authorization: require multiple approvals before agents execute critical actions safely.
• Behavioral Monitoring: detect anomalous agent behavior and trigger alerts in real time.

Who Can Use asqav?

• AI Engineers: developers building agent systems who need governance built into their stack.
• Compliance Teams: professionals responsible for ensuring AI systems meet regulatory standards.
• Enterprise Organizations: companies deploying AI agents at scale requiring audit capabilities.

Pricing

• Free Plan ($0): includes agent creation, signed actions, audit export, and tracing features.
• Business Plan (custom pricing): adds multi-party signing, SD-JWT tokens, and priority support.

Pros and Cons

Pros:

• Open Source: released under MIT license giving full transparency and community contributions.
• Framework Compatible: works with LangChain, CrewAI, LiteLLM, Haystack, and OpenAI Agents.
• Future-Proof Security: quantum-safe cryptography protects audit trails against emerging threats.

Cons:

• Python Only: currently limited to Python environments with no other language support yet.
• Early Stage: relatively new project that may lack extensive documentation and community resources.
• Setup Complexity: configuring policy rules and multi-party signing requires some technical expertise.

FAQs:

1) What frameworks does asqav support?

It works with LangChain, CrewAI, LiteLLM, Haystack, and OpenAI Agents out of the box.

2) Is asqav free to use?

Yes, the free plan includes agent creation, signed actions, audit export, and tracing.

3) What makes asqav quantum-safe?

It uses ML-DSA-65 signatures standardized under FIPS 204, designed to resist quantum attacks.

4) Does asqav help with EU AI Act compliance?

Yes, it generates compliance reports and maintains tamper-proof audit trails for regulations.

5) Can multiple people approve agent actions?

Yes, multi-party signing uses an m-of-n threshold requiring minimum approvals before execution.