einhorn-forensic-short
Pass
Audited by Gen Agent Trust Hub on Aug 31, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE]: The skill does not contain any executable scripts (Python, JavaScript, shell) or configuration files that trigger command execution. It is composed of markdown instructions and metadata for the agent to follow.
- [INDIRECT_PROMPT_INJECTION]: The instructions direct the agent to analyze primary financial filings and footnotes. This introduces a data ingestion surface where external content is processed. However, the skill lacks tool-use instructions or code execution capabilities that could be exploited by an injection in the processed data. The analysis follows a strict, evidence-based reasoning framework which acts as a natural constraint.
- [SAFE]: All external URLs referenced in the documentation (Spotify, Amazon, and Greenlight Capital) target well-known, reputable services and official organization websites. No suspicious or obfuscated links were found.
Audit Metadata