NTI LEADERBOARDNTI-1.0
Post-Quantum ML-DSA-5 Verified Benchmarks
The Standard for AI Agent Security
The definitive public benchmark for AI Agent security. Eliminating self-reported scores with offline-verifiable post-quantum security attestations.
| Rank | Framework | Score | Grade | Status | Trend | Last Scanned | Verification | |
|---|---|---|---|---|---|---|---|---|
| #1 | NTI Reference Implementation3 findings neutral-trust-infrastructure | 92 | A+ | Verified | Just now | |||
| #2 | aps-conformance-suite289 findings aps-conformance-suite | 66 | D | Failing | Just now | |||
| #3 | AutoGen641 findings autogen | 25 | F | Failing | Just now | |||
| #4 | LangChain1082 findings langchain | 16 | F | Failing | Just now | |||
| #5 | CrewAI1471 findings crewai | 0 | F | Failing | Just now |
The Reproducibility Engine
Don't trust us. Verify it yourself.
Every score on this board is backed by a cryptographic attestation. We don't guess. We parse the AST and generate offline-verifiable post-quantum attestations.
terminal — local verification
# 1. Install the official NTI static analyzer
$ pip install nti-scanner
# 2. Run local AST scan and generate cryptographic proof
$ nti-scanner scan . --submit