alterlab-rma-qualitative-coder
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection attacks as it is designed to process external, untrusted data while maintaining file-writing and search capabilities.
- Ingestion points: The skill specifically instructs the agent to read external project files, including interview transcripts, field notes, and open-ended survey responses (referenced in the 'Agentic Protocol' and 'Workflow' sections of SKILL.md).
- Boundary markers: There are no explicit instructions or delimiters (such as XML tags or Markdown blocks) defined to separate external data from system instructions, which could lead the agent to follow commands embedded within transcripts.
- Capability inventory: The skill is authorized to perform web searches for methodological guidance and has extensive file-writing capabilities, creating multiple deliverables like codebooks (
{project}-codebook-v1.md) and analysis reports ({project}-thematic-analysis.md) in the project directory. - Sanitization: Although the skill correctly mandates data de-identification for privacy purposes, it lacks mechanisms for sanitizing or escaping the content of ingested files to prevent adversarial instructions from influencing agent behavior.
Audit Metadata