The Agent Skills Directory

[PROMPT_INJECTION]: The skill's instructions create a vulnerability surface for indirect prompt injection. \n- Ingestion points: The agent is directed to ingest untrusted data from Slack, Email, Google Drive, and Calendar in the examples/3p-updates.md, examples/company-newsletter.md, and examples/faq-answers.md files. \n- Boundary markers: The skill does not define delimiters or warnings to prevent the agent from executing instructions embedded within the data it retrieves. \n- Capability inventory: The agent is granted access to read and summarize highly sensitive corporate communications and files. \n- Sanitization: There are no procedures defined for sanitizing or validating the content gathered from these external tools.

internal-comms