The Agent Skills Directory

[SAFE]: The skill contains no executable code or shell commands. All content is instructional markdown or JSON evaluation data. No obfuscation, persistence mechanisms, or privilege escalation patterns are present.
[INDIRECT_PROMPT_INJECTION]: The skill describes a workflow that ingests external data exports (CRM and billing records) for reconciliation. While this is an ingestion surface, the instructions focus on analytical accuracy and formal sign-off rather than processing untrusted instructions.
[DYNAMIC_EXECUTION]: The workflow mentions using code execution to script the bridge between data exports. This is a legitimate functional use case for data analysis agents and is performed on local data provided by the user, representing a standard feature rather than a security risk.

revenue-reporting