The Agent Skills Directory

[SAFE]: The skill is primarily educational and instructional, providing templates and strategies for business operations.
[COMMAND_EXECUTION]: The skill includes a local script, scripts/churn_impact_calculator.py, intended for revenue modeling. Analysis shows the script uses only standard Python libraries (json, sys, argparse) to perform mathematical calculations. It does not execute arbitrary shell commands or use unsafe functions like eval() or exec().
[DATA_EXFILTRATION]: No network operations or sensitive data access patterns were found. While the documentation discusses integration with payment providers like Stripe and Braintree, these are references for the user's business logic and do not involve the skill accessing credentials or making external API calls.
[PROMPT_INJECTION]: The instructions focus on establishing an expert persona for churn management. There are no attempts to bypass safety filters or override the underlying agent's core instructions.
[OBFUSCATION]: All content is provided in clear, readable text and standard code. No hidden characters, Base64 encoding of commands, or other obfuscation techniques were detected.

retention-manager