The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a Python script (scripts/check-output.py) designed to validate the output produced by the agent. This script performs basic string matching to ensure the presence of required marketing terms and checks the length of the deliverable. It operates locally and does not perform network operations or access sensitive system resources.
[INDIRECT_PROMPT_INJECTION]: The skill provides a surface for processing untrusted user data (product details, audience information) to generate a launch plan. However, the skill lacks dangerous capabilities such as network exfiltration or sensitive file writing, and the internal validation script does not process user data in a way that could lead to code execution. This represents a standard use case for text-processing agents.
[REMOTE_CODE_EXECUTION]: The instructions mention external platforms such as Product Hunt, Hacker News, and Muck Rack as part of the launch strategy. These are well-known services mentioned in a purely informational context. There are no patterns of remote script fetching or execution.

launch-planning