spark-recipe-meeting-prep
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it ingests and processes untrusted data from external sources such as email bodies and meeting transcripts. However, this is inherent to the skill's primary function and no exploitable capabilities were found.
- Ingestion points: The skill uses
spark searchto retrieve email bodies andspark meetingto retrieve transcripts inSKILL.md. - Boundary markers: There are no explicit delimiters or specific instructions provided to the agent to treat the retrieved content as untrusted data.
- Capability inventory: The skill is restricted to read-only operations using the
sparkCLI tool. It does not perform file system writes, network exfiltration, or execute arbitrary system commands. - Sanitization: No sanitization of the retrieved email or transcript text is specified, which is common for read-only summarization tasks.
Audit Metadata