Asqav signs every AI-agent action into a tamper-proof receipt (ML-DSA-65 signatures, RFC 8785 canonical JSON, RFC 3161 timestamps) that anyone can verify offline with the open-source (Apache-2.0) verifier and the Python and TypeScript SDKs.
About Asqav
Asqav is an evidence layer for AI agents. It signs each agent action into a receipt using ML-DSA-65, a post-quantum signature standard from NIST FIPS 204. The receipts are hash-chained per agent and timestamped, and can be checked offline with an open-source verifier and Python and TypeScript SDKs.
The PyPI package page, a Dify marketplace announcement and posts by the project's author on DEV Community describe the same design. They say receipts use RFC 8785 canonical JSON, RFC 3161 timestamps, and can cover blocked actions as well as allowed ones. The receipt format is described as an independent IETF Internet-Draft, not an adopted standard. Most of these details come from Asqav's own materials and partner posts, so the security and compliance claims are vendor statements, and I did not verify the cryptography myself.
Key features
- Signed receipts for agent actions
- ML-DSA-65 signatures
- Offline verification
- Python and TypeScript SDKs
- Records blocked and allowed actions
Frequently asked questions about Asqav
- What is Asqav?
- Asqav signs every AI-agent action into a tamper-proof receipt (ML-DSA-65 signatures, RFC 8785 canonical JSON, RFC 3161 timestamps) that anyone can verify offline with the open-source (Apache-2.0) verifier and the Python and TypeScript SDKs. It is listed on botshub under industry-specific agents.
- What can Asqav do?
- Key features include signed receipts for agent actions, mL-DSA-65 signatures, offline verification, python and TypeScript SDKs and records blocked and allowed actions.
- What are alternatives to Asqav?
- Other industry-specific agents on botshub include Stood, Flowmingo, Gatehouse and TapVow. See all alternatives to Asqav.