customer-research

Pass

Audited by Gen Agent Trust Hub on Aug 1, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze data from various untrusted external sources, including customer transcripts, Reddit threads, and G2 reviews. While this creates a potential surface for indirect prompt injection, the skill's instructions focus on structured extraction and synthesis, which limits the potential impact. The analysis found a lack of specific boundary markers or sanitization instructions for this ingested data, but this is consistent with the primary purpose of a research skill.
  • [EXTERNAL_DOWNLOADS]: The skill's reference materials point to established and reputable services such as Reddit, LinkedIn, and SparkToro. These references are used solely for information gathering and do not involve the execution of remote scripts or the installation of unverified software.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 1, 2026, 07:44 AM