GOV-10: Tool allow-listing and hiding
In short: Allow each agent to use only the tools on its approved list, and hide every other tool from it entirely. An agent cannot misuse a tool it cannot see. Hiding tools also reduces prompt injection risk, because injected instructions cannot call unlisted tools.
1GOV-10
What does GOV-10 require?
Allow each agent to use only the tools on its approved list, and hide every other tool from it entirely.
Why does it matter?
An agent cannot misuse a tool it cannot see. Hiding tools also reduces prompt injection risk, because injected instructions cannot call unlisted tools.
How do you test it?
- List the tools an agent can see at runtime.
- Compare that list with its approved tool list.
- Ask the agent to call an unlisted tool and confirm it cannot.
Which platforms document GOV-10?
Documented by NewCore and Aembit.
- 9.10NewCore documentedSource: NewCore: Govern, read
- 6.55Aembit documentedSource: Aembit IAM for Agentic AI is generally available, read
- 6.85Descope partly documentedSource: Descope Agentic Identity Hub, read
- 8.35Silverfort not found in public materialChecked: Silverfort: AI Agent Security; Silverfort unveils AI Agent Security; Silverfort and Google Cloud Agent Gateway; Silverfort pricing, read
- 7.30Astrix Security not found in public materialChecked: Astrix at RSAC 2026; Astrix product; Cisco and Astrix Security, read
- 7.30Cyera Identity not found in public materialChecked: Oasis Agentic Access Management; Cyera completes Oasis Security acquisition, read
- 6.55Token Security not found in public materialChecked: Token Security home, read
- 6.55SailPoint Entro not found in public materialChecked: SailPoint agentic AI security; SailPoint closes Entro Security acquisition, read
- 6.35Clutch Security not found in public materialChecked: Clutch Security platform, read
- 5.50Idira not found in public materialChecked: Palo Alto Networks introduces Idira; CyberArk Secure AI architecture docs; Palo Alto Networks completes CyberArk acquisition, read
Which framework references apply?
- OWASP Agentic Top 10ASI02 Tool Misuse: Agents bending legitimate tools into harmful outputs. Source: OWASP Top 10 for Agentic Applications 2026
Which controls are related?
- GOV-06Per-action policy
- GOV-17MCP server registry
- GOV-22Data access boundaries
Glossary: tool allow-list, mcp.
Reviewed by The Register desk