Threat Model
AnchorWorks is designed to address retroactive record modification, unverifiable governance claims, and timing disputes. It does not address inadequate governance or compromised host systems.
Addressed Threats
Retroactive modification of governance records. Claims of governance without evidence. Disputes about timing of governance events.
Not Addressed Threats
Inadequate governance declarations. Malicious operators who control both AI and adapter. Side-channel attacks on the host system.
Trust Assumptions
AnchorWorks assumes the host system is not fully compromised and that ForgeRun operates correctly.
Understanding what threats AnchorWorks addresses and what it does not address.
Addressed Threats
Retroactive modification of governance records. Claims of governance without evidence. Disputes about timing of governance events.
Not Addressed Threats
Inadequate governance declarations. Malicious operators who control both AI and adapter. Side-channel attacks on the host system.
Trust Assumptions
AnchorWorks assumes the host system is not fully compromised and that ForgeRun operates correctly.