fiftyone-sdk-guidance

Pass

Audited by Gen Agent Trust Hub on Jul 29, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill uses a public, read-only MCP server (https://voxel51.mcp.kapa.ai) for documentation search. This is a well-known service provided by the vendor (Voxel51) and does not require sensitive API keys from the user.
  • [COMMAND_EXECUTION]: The README and SKILL.md contain instructions to run claude mcp add to install the documentation tool. These commands are benign, require explicit user confirmation before execution in the agent environment, and are a standard part of the platform's extensibility.
  • [EXTERNAL_DOWNLOADS]: The skill references standard documentation and installation paths from docs.voxel51.com and skil.sh, which are official vendor resources.
  • [PROMPT_INJECTION]: The skill includes instructions to ensure the agent uses semantic search correctly (e.g., using full sentences instead of keywords). These are positive quality-of-service instructions and do not attempt to bypass safety filters or override system behavior maliciously.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 29, 2026, 02:35 AM
Security Audit — agent-trust-hub — fiftyone-sdk-guidance