analyze-chart
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns such as prompt injection, data exfiltration, or obfuscation were detected.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources (e.g., customer feedback via Amplitude:get_feedback_insights). While this creates a theoretical surface for indirect prompt injection, the skill's capabilities are limited to data retrieval and report generation without dangerous actions like file writing or command execution.
- Ingestion points: Amplitude:get_feedback_insights and Amplitude:get_feedback_mentions in SKILL.md
- Boundary markers: Absent
- Capability inventory: Data retrieval tools only; no code execution or file system writes
- Sanitization: Not explicitly specified in the skill instructions
Audit Metadata