The Agent Skills Directory

Data Retrieval and Processing: The skill uses the ask_work_iq tool to access user profile details (email, display name) and Teams channel content. This data is used to personalize and generate the digest. This is a core part of the skill's functionality and is conducted through sanctioned internal mechanisms.
Processing of Untrusted Content: In Steps 3 and 5, the skill retrieves message content from Teams channels. Because this content originates from various users, it could theoretically contain text designed to influence the model's summarization behavior (indirect prompt injection). This is a standard consideration for any tool that processes external communication data, and the risk is mitigated by the skill's specific focus on categorization and summarization.

channel-digest