helix-query-from-gremlin

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to translate Gremlin traversals (untrusted input) into Rust SDK code. If the input Gremlin query contains embedded malicious instructions, the agent might follow them during the translation process. The skill provides capability for the agent to execute tools like helix-mcp to inspect live databases.
  • Ingestion points: Processes Gremlin queries, TinkerPop traversals, and results from helix-mcp (explicitly noted as untrusted data in the text).
  • Boundary markers: Absent. There are no specific delimiters defined for the input queries to separate them from instructions.
  • Capability inventory: The skill is intended to generate code for client.query(request). While it doesn't explicitly call shell commands, it interacts with local repository files (docs/, examples/) and helix-mcp to inform its output.
  • Sanitization: The instructions advise to treat helix-mcp results as untrusted, but there is no specific guidance on sanitizing the input Gremlin strings themselves before translation.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 03:02 PM
Security Audit — agent-trust-hub — helix-query-from-gremlin