running-flow

Pass

Audited by Gen Agent Trust Hub on May 2, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the robomotion and robomotion-deskbot CLI tools. These tools are used locally to validate automation scripts, connect to robot sessions, and trigger flow executions. This is standard functionality for the vendor's automation platform.
  • [DATA_EXFILTRATION]: The skill reads session log files from ~/.config/robomotion/agent/logs/sessions/ (Linux/macOS) and %LOCALAPPDATA%\Robomotion\agent\logs\sessions\ (Windows). These files are specific to the vendor's toolchain and are used by the agent to observe execution status and errors.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8) by design.
  • Ingestion points: The agent reads and processes untrusted JSONL session logs from the local filesystem (SKILL.md).
  • Boundary markers: None identified. The agent is instructed to parse raw log events to drive a 'fix' loop.
  • Capability inventory: The agent can execute shell commands (robomotion run), modify source code (main.ts), and manage environment variables.
  • Sanitization: There is no evidence of sanitization or validation of the log content before it is processed by the agent to determine its next actions.
Audit Metadata
Risk Level
SAFE
Analyzed
May 2, 2026, 01:26 PM
Security Audit — agent-trust-hub — running-flow