The Agent Skills Directory

[Indirect Prompt Injection Surface]: The skill is designed to ingest and process significant amounts of external data, including email subjects, chat messages, and calendar entries. This creates a potential surface where instructions hidden within these messages could influence the agent's behavior.
Ingestion points: Processes data from Microsoft 365 via m365_* tools, specifically reading emails (Inbox and Sent), chat messages, channel posts, and calendar subjects (identified in SKILL.md).
Boundary markers: The skill lacks explicit instructions for the agent to ignore or delimit instructions that might be embedded within the processed email or chat content.
Capability inventory: The agent has the capability to read extensive personal and organizational communications across multiple platforms (Email, Teams, Calendar).
Sanitization: There are no specific technical sanitization or validation steps mentioned for the retrieved content, although the skill includes instructions for the agent to anonymize output and avoid private content.

pattern-radar