economics
Pass
Audited by Gen Agent Trust Hub on Mar 8, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [SAFE]: The skill's instructions are focused on legitimate financial analysis and do not contain prompt injection attempts, obfuscation, or instructions to bypass safety guidelines.
- [COMMAND_EXECUTION]: Uses the standard grep utility to search for business-related constants like pricing and credits within the local project source code. This is an appropriate use of system tools for the task of configuration discovery.
- [DATA_EXFILTRATION]: While the skill reads project files to estimate token usage and costs, it specifically targets business constants and prompt templates rather than sensitive system files, private keys, or credentials. No data is transmitted to untrusted external domains.
- [EXTERNAL_DOWNLOADS]: Suggests performing web searches for pricing on the official websites of well-known providers such as OpenAI, Anthropic, and Google. These are trusted sources and the activity involves informational lookup rather than the execution of remote scripts.
Audit Metadata