opportunity-scoring-review
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes user-supplied opportunity descriptions which constitute untrusted data, creating a surface for indirect prompt injection where malicious instructions could attempt to bias the scoring outcome.
- Ingestion points: Product, growth, and strategic opportunities provided by the user for scoring (SKILL.md).
- Boundary markers: Absent. The instructions do not define delimiters or provide clear guidelines for the agent to treat input purely as data.
- Capability inventory: The skill's capabilities are restricted to conversational analysis and text generation; no system-level or network tools are utilized.
- Sanitization: No explicit validation or filtering of input content is specified in the prioritization workflow.
- Mitigation: Wrap external opportunity data in specific delimiters (e.g., triple backticks or XML tags) and add a rule to ignore any instructions embedded within the opportunity descriptions.
Audit Metadata