The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from external server resources while maintaining capabilities to perform state-changing operations.\n
Ingestion points: External data is ingested through various list tools, including list_websites, list_ssls, and list_installed_apps, as documented in references/tools.md.\n
Boundary markers: Clear instructions in references/operation-contract.md direct the agent to treat all tool results and imported documents strictly as data and never as instructions.\n
Capability inventory: The skill possesses write capabilities such as create_website, create_ssl, and create_database as defined in references/tools.md.\n
Sanitization: The skill does not implement programmatic sanitization, instead relying on instructional prompts to prevent the execution of instructions embedded in retrieved data.

1panel-system