skills/no-phux/skills/using-phux-mcp/Gen Agent Trust Hub

using-phux-mcp

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill interacts with terminal output and persistent shell state which constitutes an attack surface for indirect prompt injection.
  • Ingestion points: Untrusted data may enter the agent context through tools that read terminal state such as phux_snapshot, phux_status, and phux_doctor (SKILL.md).
  • Boundary markers: The instructions do not specify explicit delimiters or "ignore embedded instructions" warnings for data read from the terminal.
  • Capability inventory: The skill possesses capabilities to execute commands and send keys through phux_run, phux_send_keys, and phux_agent_prompt (SKILL.md).
  • Sanitization: There is no mention of sanitization or filtering of the terminal output before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 07:38 PM
Security Audit — agent-trust-hub — using-phux-mcp