pull-call-review
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data in the form of sales call transcripts provided by the user. There are no explicit instructions for the agent to sanitize the input or ignore instructions embedded within these transcripts, which creates a potential surface for manipulating the agent's behavior.
- Ingestion points: User-provided sales call transcripts mentioned in the SKILL.md description and instructions.
- Boundary markers: Absent; no specific delimiters or "ignore embedded instructions" warnings are defined in the instructions.
- Capability inventory: The skill is primarily designed for generating text-based analysis and does not explicitly request network or file-system tool access.
- Sanitization: No sanitization, escaping, or validation routines are specified for the transcript content.
Audit Metadata