vectordb
Pass
Audited by Gen Agent Trust Hub on Mar 13, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill implements a Retrieval-Augmented Generation (RAG) pattern in the RAGService class that is susceptible to indirect prompt injection.
- Ingestion points: Document content provided via add_document is stored and later retrieved to be placed into the agent's prompt context.
- Boundary markers: The prompt template in the query method uses weak headers (Context: and Question:) to separate document content from user questions, lacking robust delimiters or defensive instructions.
- Capability inventory: The skill is configured with broad capabilities including Bash, Read, Write, and Edit tools, which could be exploited if an injection is successful.
- Sanitization: No input sanitization or validation of the ingested document content is performed before it is interpolated into the prompt.
Audit Metadata