connect

Pass

Audited by Gen Agent Trust Hub on Sep 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local commands to determine the version of the coding agent (claude --version or codex --version).
  • [COMMAND_EXECUTION]: It employs a specific command-line utility, altitude, to perform the device pairing sequence (altitude connect) and verify the connection state (altitude status).
  • [INDIRECT_PROMPT_INJECTION]: The agent is instructed to process and display output from the altitude tool, specifically verification URLs and codes. While this is a standard part of the pairing process, it involves the agent handling external data from tool responses.
  • Ingestion points: Output from the altitude connect command in SKILL.md.
  • Boundary markers: None explicitly defined for the tool output.
  • Capability inventory: Executes bash commands via altitude and version check tools as defined in the allowed-tools frontmatter.
  • Sanitization: Not explicitly defined; the skill relies on the expected output format of the altitude tool.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 6, 2026, 04:58 AM
Security Audit — agent-trust-hub — connect