Security & Infrastructure Analysis
The Singulant Research
Technical analysis of AI-agent authorization, Web3 infrastructure, cryptographic verification, identity systems, deterministic recovery, and real-world security failures.
Editorial standards
Every publication is dated and separates what we checked from what others report:
- Independently verified facts: claims we checked against primary sources or public chain data, with the date of the check.
- Third-party reported findings: what the cited source says, attributed and not presented as confirmed.
- Technical interpretation: our analysis, labeled as such.
- AI⁴ design implications: how a finding may inform AI⁴ design. These describe direction under development or evaluation, not deployed controls.
- Unresolved questions and limitations: what we could not verify.
The Singulant Research is the research section of thesingulant.ai, not an independent institution. It does not publish market or token commentary.
Categories
Threat Intelligence
Documented attacks, trust-boundary failures, deceptive resolution, identity spoofing, and related incidents.
1 publicationArchitecture & Verification
AI⁴ design principles, independently verified engineering evidence, authorization models, delegation, and recovery.
No publications yetInfrastructure Watch
Material developments in Web3 naming, interoperability, agent infrastructure, ENS, Unstoppable Domains, and related protocols.
No publications yet
Publications
-
Threat Intelligence
When a Web3 Domain Resolves Correctly but Sends Funds to the Wrong Person
A reported scam used an Unstoppable Domains name built from an exchange deposit address. According to the source, the wallet resolved it as designed, to an address the sender did not intend to pay. Why a valid blockchain address is not proof of an authorized destination.
Primary source: Habr (sansmaster)