The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources which could potentially contain malicious instructions meant to influence the agent's behavior during report generation.
Ingestion points: Data is retrieved from ~~бухгалтерия (accounting), ~~платежи (payments), and ~~crm in Steps 1 and 2.
Boundary markers: The skill does not define specific delimiters or use instructions to disregard commands embedded within the fetched business data.
Capability inventory: The skill utilizes Bash, Read, and WebFetch tools, providing a surface for command execution or network activity if an injection occurs.
Sanitization: There is no explicit logic to sanitize, escape, or validate the content of the financial and CRM data before it is interpolated into the narrative generation prompt.
[COMMAND_EXECUTION]: The skill requires the Bash tool to perform the export of the final report in Step 6. While this is a functional requirement for generating PDF files or ASCII tables, the execution of shell commands based on aggregated external data presents a theoretical risk of command injection if the input data is not properly constrained.

quarterly-review