crash-out
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill contains no malicious patterns, prompt injections, or unauthorized data access. Its instructions are focused on a "crash-out" reasoning process where the model attempts to disprove its own initial draft to ensure accuracy.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow that ingests external data from web searches and documentation to verify factual claims. While this introduces an ingestion surface for untrusted data, it is necessary for the skill's primary purpose of verification and utilizes the agent's existing tool-use guardrails.
- [DYNAMIC_EXECUTION]: The instructions direct the agent to run code or commands to validate technical statements. This behavior is a standard use of an agent's execution environment for factual validation and does not introduce new or unexpected execution risks.
Audit Metadata