Armor scores each tool call against a set of rules plus a slice of the transcript, prioritizing user intent. If the user asked "Push our work", a (bash tool: git push) is allowed, while a dangerous command the user never asked for is flagged.
Armor scores agent tool calls against rules and user intent
AISummary
OpenRouter says Armor checks each tool call against a set of rules and a slice of the transcript, prioritizing user intent. In its example, a git push is allowed when the user asked to push their work, while a dangerous command the user never requested is flagged.
Post on XView on X
Source: OpenRouter · x.comPublished