qualitative-research
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [Indirect Prompt Injection] (LOW): The skill is designed to ingest and analyze external research data using the provided toolset.\n
- Ingestion points: Research data, transcripts, and field notes used for thematic analysis as defined in the protocol and tool definitions.\n
- Boundary markers: Absent. The instructions do not define specific delimiters or guidelines to distinguish between processed data and system instructions.\n
- Capability inventory: The skill possesses the WebSearch, WebFetch, Read, Grep, and Glob tools.\n
- Sanitization: Absent. No explicit sanitization or escaping mechanisms for external content are described in the instructions.\n- [Prompt Injection] (SAFE): No evidence of direct prompt injection, system prompt extraction, or safety filter bypass attempts was found. Natural instructional language is used appropriately for the researcher persona.\n- [Data Exposure & Exfiltration] (SAFE): No hardcoded secrets, API keys, or sensitive local file paths (e.g., .ssh, .aws) were identified.\n- [External Downloads & RCE] (SAFE): No external package installations or piped remote execution patterns (e.g., curl to bash) were detected within the skill content.
Audit Metadata