skills/product-on-purpose/thinking-framework-skills/think-evidence-vs-inference-sort/Gen Agent Trust Hub
think-evidence-vs-inference-sort
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions and metadata are focused on a reasoning methodology. No malicious code, obfuscated content, or unauthorized system interactions were found. The skill does not use any tools or external scripts.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied text (documents or conclusions), which represents an ingestion surface. Ingestion points: Text prompts or documents provided by the user for sorting in
SKILL.md. Boundary markers: The skill requires the use of a specific markdown template (references/TEMPLATE.md) to structure findings, providing a logical boundary. Capability inventory: The skill possesses no capabilities for network operations, file writing, or command execution. Sanitization: While no explicit sanitization is defined, the skill is designed to treat input as data to be analyzed rather than instructions to be followed, which prevents typical prompt injection outcomes.
Audit Metadata