customer-research
Pass
Audited by Gen Agent Trust Hub on Aug 28, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely composed of documentation and YAML configuration. There are no scripts, binaries, or executable components included.- [INDIRECT_PROMPT_INJECTION]: The skill defines a process for ingesting and analyzing untrusted external data, such as public forum posts and customer reviews. While this creates a surface for indirect prompt injection, the skill includes instructions to separate observation from interpretation and to verify synthesized summaries against raw material, which are effective manual safety practices for this use case.
- Ingestion points: Ingests transcripts, surveys, reviews, and public community signals as described in SKILL.md and references/research-modes.md.
- Boundary markers: No explicit delimiter usage or specific 'ignore instructions' markers are defined.
- Capability inventory: The skill does not define or request tools for command execution, file system modification, or network requests.
- Sanitization: No automated sanitization logic is present in the skill instructions.- [DATA_EXFILTRATION]: The operating contract in SKILL.md explicitly prohibits contacting participants, scraping restricted spaces, or exposing sensitive data without authorization.
Audit Metadata