data-engineering-data-driven-feature
Pass
Audited by Gen Agent Trust Hub on Apr 1, 2026
Risk Level: SAFENO_CODEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill instructions and accompanying resource documentation describe a legitimate process for building and testing product features based on data insights. No malicious commands, unauthorized data access, credential exposure, or obfuscated contents were identified.
- [NO_CODE]: The skill is composed entirely of Markdown documentation and prompt instructions. It does not ship any executable scripts, binary files, or external software dependencies.
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it interpolates user-supplied feature descriptions ($ARGUMENTS) into prompts for multiple subagents. Ingestion points: SKILL.md (found in multiple subagent task prompts). Boundary markers: Absent. Capability inventory: The skill uses the Task tool to delegate complex operations to various subagent roles. Sanitization: None identified. The skill relies on the safety guardrails of the underlying large language models.
Audit Metadata