The Agent Skills Directory

[SAFE]: The skill implements a structured workflow for financial data analysis. It includes safety constraints such as 'Approval gates' that explicitly forbid the agent from recommending specific prices or making changes to connected accounting systems.
[COMMAND_EXECUTION]: While the skill lists Bash as an allowed tool, it does not contain any specific shell commands or scripts. It uses the tool only as a scoped execution environment for the agent to process data and perform calculations as needed by the workflow.
[DATA_EXPOSURE]: The skill describes ingesting data from accounting systems (1C, MojSklad) or CSV files. This is consistent with its stated purpose of margin analysis. There are no patterns suggesting data exfiltration or unauthorized access to sensitive system files.
[INDIRECT_PROMPT_INJECTION]: The skill has an ingestion surface as it processes external data from accounting software and payment gateways. However, it does not pass this data into unsafe sinks (like eval or shell interpolation) and contains instructions for the agent to maintain objective data reporting without automated decision-making.

price-check