AgentDB Vector Search

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill implements a Retrieval-Augmented Generation (RAG) pattern, creating a surface for indirect prompt injection.
  • Ingestion points: Data is stored in the database via methods such as insertPattern, storeWithEmbedding, and batchStore defined in SKILL.md.
  • Boundary markers: The provided example for building RAG pipelines lacks explicit delimiters or instructions to isolate retrieved context from the agent's system prompt.
  • Capability inventory: The skill performs vector database operations including similarity search, hybrid search, and batch storage.
  • Sanitization: No sanitization or validation of the retrieved content is demonstrated before the data is used to construct prompts.
  • [EXTERNAL_DOWNLOADS]: The skill facilitates the download and execution of the agentdb package via npx. This is a vendor-owned resource provided by the author 'ruvnet'.
  • [COMMAND_EXECUTION]: The skill utilizes several CLI commands for initializing the vector database, performing semantic queries, and conducting performance benchmarks.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 04:33 PM