How to meet it
OWASP AISVS v1.01
Model Context Protocol (MCP) Security · Schema, Message, and Input Validation
Screen MCP responses for prompt injection
Verify that MCP tools/list and tools/call responses are screened for indirect prompt injection before being injected into the model context.
OWASP AISVS v1.01, requirement 10.4.2
In plain English
Screen tool descriptions and results for hidden instructions before the model sees them. A malicious server can put instructions in either.
What good looks like
- Tool descriptions and results are run through injection detection.
- Flagged content is blocked.
- Events are logged.
How teams usually do it
- The same injection checks used for user input (see 2.1.3), applied in the MCP client.
Evidence an auditor would accept
- Configuration and tests with poisoned tool descriptions.
Common mistakes
- Screening tool results but trusting tool descriptions.
Not sure which requirements apply to your system? Take the assessment.
Requirement text from the OWASP AI Security Verification Standard, licensed under CC BY-SA 4.0. This guidance is shared under the same licence.