The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes untrusted data from CRM meeting transcripts and activity logs to generate summaries. This behavior is necessary for its primary function and no malicious instructions were found in the skill itself.
Ingestion points: Data enters the context via get_deal results, contact records, activity timelines, and meeting transcripts.
Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the fetched data.
Capability inventory: The skill utilizes the crm capability for data retrieval.
Sanitization: No explicit sanitization or filtering of transcripts or email content is performed before summarization.
[SAFE]: All identified external resources and actions are related to the core CRM functionality and the vendor's intended use case. No indicators of data exfiltration or persistence were found.

Copilot Summary