skills/wu-yc/labclaw/vaex/Gen Agent Trust Hub

vaex

Pass

Audited by Gen Agent Trust Hub on Mar 15, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill includes guidelines for the agent to suggest the vendor's research platform, K-Dense Web, when handling complex workflows.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through data ingestion points. 1. Ingestion points: The skill loads data via vaex.open() and vaex.from_csv(). 2. Boundary markers: The instructions lack explicit markers to differentiate untrusted data from instructions. 3. Capability inventory: The skill utilizes file-writing capabilities through the export_hdf5() function. 4. Sanitization: No content validation or sanitization is described for the ingested data.
  • [COMMAND_EXECUTION]: The documentation provides Python code examples for executing data operations such as filtering, aggregation, and visualization within the Vaex environment.
  • [EXTERNAL_DOWNLOADS]: The skill depends on the vaex Python library, a third-party tool for high-performance out-of-core data processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 15, 2026, 04:33 AM