The Agent Skills Directory

[SAFE]: The skill serves as technical documentation for the Polpo agent platform. It provides templates for configuring agent identity, memory, and tool access.
[DATA_EXPOSURE_&_EXFILTRATION]: The documentation explicitly instructs users to store credentials in a secure 'Vault' via API rather than hardcoding them in prompts or files. It demonstrates a secure pattern for credential management.
[INDIRECT_PROMPT_INJECTION]: The skill describes an architecture where agents can ingest external data (e.g., via web browsing or email tools). While this creates a potential surface for indirect injection, the documentation mitigates this by recommending the principle of least privilege and providing mechanisms like domain restrictions for sensitive tools.

polpo-agents