mercury-shared
Pass
Audited by Gen Agent Trust Hub on Jun 13, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to use the
mercuryCLI for financial data retrieval and thedatecommand for timestamp anchoring. These are legitimate uses of vendor-provided and system tools.- [DATA_EXPOSURE]: While the skill interacts with sensitive financial data, it includes explicit security guidelines to mask account and card numbers (showing only the last 4 digits) before displaying information to the user.- [INDIRECT_PROMPT_INJECTION]: The skill defines patterns for ingesting external CLI data. It mitigates potential risks by recommending structuredjsonlformats and specific field filtering via GJSON transformations to limit the data brought into the agent's context.- [EXTERNAL_DOWNLOADS]: References the official GJSON syntax documentation on GitHub, which is a well-known service used for technical reference.
Audit Metadata