ddd4j-boot-cache
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's primary function involves reading and analyzing external project artifacts such as POM files, source code, and configuration properties.\n
- Ingestion points: The workflow in
SKILL.mdexplicitly directs the agent to gather evidence from the target project's POMs, source code, and tests.\n - Boundary markers: The instructions do not define specific delimiters or boundary markers to isolate these untrusted inputs from the agent's internal reasoning, increasing the risk that malicious content in the analyzed files could manipulate the agent's behavior.\n
- Capability inventory: The skill includes capabilities to read files and execute tests (as per
Workflow Step 4), which provides an execution path for malicious code embedded in the project being analyzed.\n - Sanitization: No sanitization or validation logic is specified for the ingested content, although the skill does advise redacting sensitive values from the final output.\n- [COMMAND_EXECUTION]: The documentation and examples (e.g.,
SKILL.mdandexamples/04-cache-spi-change-refusal.md) suggest the use ofnpx skills addto install related skill modules. This involves executing a command that pulls and runs code from a package registry, representing a potential execution vector for external code.
Audit Metadata