adelic-ai

Pinned Loading

1. omega omega Public

   A research instrument working toward one map of detection knowledge — the major rule libraries, where they intersect and where they don't, resolved against MITRE ATT&CK.

   Python 1

2. sift sift Public

   Sensitive-data discovery over files — finds PII and secrets, controls false positives with Benjamini–Hochberg FDR (per document), and reports the files it could not read.

   Python

3. agentwatch agentwatch Public

   EDR-style detection for LLM coding agents — reconciles what an agent said it did (transcript) against what it actually did (OS audit) to flag actions with no authorizing intent. Also reconciles Kub…

   Python

4. warrant warrant Public

   Scoped, auditable authorization for AI agents acting on behalf of humans — SPIFFE identity, OAuth token exchange, Cedar policy.

   Python

5. siphonophore siphonophore Public

   An SDK for binding delegated agent authority to mediated, independently attributable execution, with a reference harness demonstrating the architecture.

   Python

6. canon canon Public

   A research substrate for justified computation: machine-checkable results with evidence, provenance, validity, custody, and explicit guarantees.

   Python