application-security-testing

Pass

Audited by Gen Agent Trust Hub on Sep 2, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill guides the agent to process data from external, untrusted sources including source code repositories, live web application responses, and API schema definitions. This ingestion process creates a surface for indirect prompt injection, where malicious instructions embedded in the analyzed assets could attempt to influence the agent's behavior or reporting.
  • Ingestion points: Data is ingested from repositories (monorepos or services), live web/staging environments, and API schemas (REST/GraphQL/gRPC).
  • Boundary markers: The orchestration instructions do not define specific delimiters or instructions to ignore embedded commands within the processed data.
  • Capability inventory: The skill coordinates the use of multiple sub-skills that perform file system reads, network operations, and CLI tool execution.
  • Sanitization: No specific sanitization or filtering logic is implemented at this orchestration level to handle potentially malicious content from the assets being tested.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 2, 2026, 02:08 AM
Security Audit — agent-trust-hub — application-security-testing