skills/product-on-purpose/thinking-framework-skills/think-ladder-of-inference-check/Gen Agent Trust Hub
think-ladder-of-inference-check
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists entirely of markdown instructions, templates, and evaluation cases. No signs of prompt injection, data exfiltration, obfuscation, or unauthorized command execution were found in the skill's instructions, examples, or metadata. It does not ship with any executable scripts or binaries.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied data (conclusions and observations). While this represents a surface for indirect prompt injection, the risk is negligible as the skill lacks high-risk capabilities such as network access, file system writes, or subprocess execution. The mandatory use of a structured markdown table template for output further reduces the risk of the agent following instructions embedded in the analyzed data.
- Ingestion points: SKILL.md instructions for processing user conclusions and data.
- Boundary markers: The output is constrained by the references/TEMPLATE.md structure.
- Capability inventory: None; the skill does not invoke any tools or execute code.
- Sanitization: None specified, but the constrained output format acts as a functional filter.
Audit Metadata