spark-recipe-weekly-digest
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
sparkCLI tool to retrieve information from the user's email, calendar, and team assignments. These commands are consistent with the skill's stated purpose and the requireduse-sparkbase skill. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources, specifically email content and meeting transcripts. 1. Ingestion points: Step 2 (email filtering) and Step 4 (meeting transcripts) in SKILL.md. 2. Boundary markers: Not present. 3. Capability inventory: Read-only access to communication and calendar data via the
sparktool. 4. Sanitization: Not present.
Audit Metadata