spark-persona-team-lead
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides a persona template for the Spark CLI tool (authored by Readdle). It uses standard commands such as
spark team,spark emails, andspark actionto manage team workloads and shared inboxes. - [INDIRECT_PROMPT_INJECTION]: The skill involves processing external data by reading email threads and shared inbox content, which represents an indirect prompt injection surface.
- Ingestion points: Email thread content is retrieved via the
spark thread <id>command as described inSKILL.md. - Boundary markers: No explicit delimiters or boundary markers for external data are specified in this persona skill.
- Capability inventory: The skill utilizes capabilities to assign emails, mark threads as done, and post comments using
spark actionandspark comment. - Sanitization: No specific sanitization or filtering logic is defined for the content retrieved from email threads.
- [SAFE]: All operations are consistent with the intended use of the Spark email management suite. No remote code execution, obfuscation, or persistence mechanisms were detected. The example emails (e.g., alice@co.com) and dates are used for illustrative purposes and do not pose a risk.
Audit Metadata