nanomind

opena2a-org

On-device models that detect and classify AI agent attacks. A 2M-param classifier (<1ms, offline) and a 1.7B-param analyst, shipping inside HackMyAgent, Secretless, and the OpenA2A CLI.

Last commit

1mo ago

Branch

chore/contact-email-to-info

Language

Python

Stars

Commits (30d)

Open issues

Open PRs

Working tree

clean

Linked todos (0)

No todos linked to this repo.

Recent commits

chore: point published contact email to info@opena2a.org

1mo ago

docs(analyst): add MLX Apple-Silicon GPU path; correct GGUF Metal guidance (#40)

1mo ago

docs(analyst): sync MODEL-CARD.md to the published HF card restructure (#39)

1mo ago

docs(analyst): correct fp-suppression to ~77% and gate to threshold 0.90 (#38)

1mo ago

feat(daemon): add explicit classify/abstain signal to /v1/infer (#131) (#36) (#37)

1mo ago

feat(daemon): add explicit classify/abstain signal to /v1/infer (#131) (#36)

1mo ago

harden(daemon): reject symlink at artifact-dir boot verify; retry healthz probe (#35)

1mo ago

nanomind-analyst 0.1.3: ship input-classifier at gate 0.90 + harden the upgrade path (#34)

2mo ago

fix(analyst): make boot healthz probe threshold-independent (#33)

2mo ago