Unverified50% confidenceFactExact time
In the demo, disabled tools are completely invisible to the agent rather than simply having their calls rejected.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/28/2026
First Seen
Valid until: 1/26/2027
Sources
AI Agent Control in Practice: A Guardrail Demo for MCP Servers
youtubeSafeguard10/28/2026
Related Entities
Related Claims
UnverifiedAgenthof's demonstration does not prove isolation: the isolation pod is a host process launched by a test double, and nothing in this run proves that the Agent cannot access the provider on its own; for this the project maintains a separate Podman-based proof, including a真正的 rootless container, --network none, and a check for no egress.66% similarUnverifiedThe plugin allows the creation of clean Agents with no preset context, building capability combinations from scratch.63% similarUnverifiedA tool is only useful when it is correctly configured and its credentials permit the operation; if the tool's credentials do not allow a certain operation, the Agent cannot authorize the use of that tool.61% similarUnverifiedPrompt injection cannot be fully filtered, only contained, because it is hidden instructions buried in the content an Agent reads, and the model treats them like user instructions.60% similarUnverifiedAn agent that is completely idle (calling no tools and outputting no text at all) can pass all safety test cases that only check 'does not call tools' without checking reply content.57% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1107607API
curl https://kongchang.com/api/v1/knowledge/claims/1107607MCP
get_claim(id=1107607)