The Agent Skills Directory

[SAFE]: The skill provides documentation and instruction for using the gog CLI tool to manage Google AdSense data. It emphasizes safety practices such as using --readonly for data retrieval, --dry-run for verifying mutations, and explicit account selection to prevent accidental actions on incorrect accounts.
[INDIRECT_PROMPT_INJECTION]: The skill identifies a surface for indirect prompt injection when reading content from Google AdSense sites or reports. It provides a specific safety mitigation by recommending the use of the --wrap-untrusted flag, which helps the agent distinguish between tool instructions and external data.

gog-adsense