painpoint-validate
Pass
Audited by Gen Agent Trust Hub on Jun 22, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill is composed entirely of instructional markdown and schema definitions. It defines a rigorous evidence-based process for product validation without the use of executable scripts or sensitive system access.
- [NO_CODE]: There are no executable files (Python, JavaScript, shell scripts) or external software dependencies associated with this skill. It operates solely through natural language instructions and structured data templates.
- [DATA_EXFILTRATION]: While the skill involves web searching and reading project files, these actions are scoped to the primary purpose of research and reporting. Data is written to local project files (e.g., 04-painpoints.md) rather than transmitted to unauthorized external endpoints.
- [PROMPT_INJECTION]: The skill uses structured 'Evidence Blocks' and 'Research Plans' which serve as organizational boundaries. These structures encourage the model to treat external data as facts for analysis rather than as authoritative instructions, reducing the risk of indirect prompt injection.
Audit Metadata