HN Showcase
Back to browse
GitHub Repository

Spec + offline verifier for Manifest-InX Evidence Bundles (EBS) v0.1: capture/pin artifacts, replay deterministically offline, reject drift/tamper.

2 starsPython

Make AI and automation pipelines fail-closed

by oneinx·Mar 8, 2026·1 point·1 comment
Visit ProjectView on HN

AI Analysis

●●SolidBig BrainZero to One

Deterministic offline verification of AI pipeline outputs with Merkle hashing—novel framing, early stage.

Strengths
  • •Solves real determinism problem in AI-driven infrastructure: nondeterministic provider calls captured at boundary
  • •Spec-first design with offline verifier shipped before runtime, enabling independent auditing
  • •Threat model and capture boundary clearly documented, not hand-waved
Weaknesses
  • •Gateway runtime not yet public; unclear how bundles are emitted and integrated into real CI/CD
  • •Early v0.1; no production users mentioned, and proof kit is example-only
Target Audience

DevOps engineers, platform teams deploying AI agents, compliance-heavy organizations

Similar Projects

AI/ML●●●Banger

97% on SWE-bench Verified with subscription-token agents

97% on SWE-bench Verified with full artifact transparency, not just a score claim.

Big BrainZero to One
kimjune01
2021d ago