701-technologies-openapi
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill mandates the agent to execute local shell commands for project verification, specifically
./mvnw compile,mvn compile,./mvnw clean verify, andmvn clean verify. While standard for Java development, these commands execute the local build system which could be leveraged if the project's build configuration is compromised. - [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to read, review, and modify OpenAPI specifications, which represent an untrusted data ingestion surface. Maliciously crafted specifications could contain instructions designed to override the agent's behavior.
- Ingestion points: OpenAPI files (YAML/JSON) read during the assessment phase in
SKILL.mdandreferences/701-technologies-openapi.md. - Boundary markers: The skill does not provide specific instructions or delimiters to the agent to ignore or escape embedded instructions within the API documentation fields (e.g., descriptions, examples).
- Capability inventory: The agent has the ability to execute shell commands (Maven) and write to the file system.
- Sanitization: No sanitization or validation logic is defined to filter natural language instructions out of the structured OpenAPI data before processing.
Audit Metadata