python-executor

Pass

Audited by Gen Agent Trust Hub on Sep 7, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill is designed to execute arbitrary Python code on a remote sandboxed infrastructure provided by inference.sh.
  • [COMMAND_EXECUTION]: Utilizes the belt CLI tool to facilitate the execution of Python scripts as part of its primary functionality.
  • [INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection because it accepts and executes Python code provided as text input, which could be influenced by untrusted external data.
  • Ingestion points: The code input parameter defined in the schema and used in code examples (SKILL.md).
  • Boundary markers: None present to distinguish instructions from potentially malicious data embedded in the code string.
  • Capability inventory: The skill possesses the capability to execute shell commands via the belt CLI and run arbitrary Python code in a remote environment.
  • Sanitization: No evidence of sanitization, validation, or escaping of the provided Python code before it is passed to the execution environment.
  • [EXTERNAL_DOWNLOADS]: The documentation references external resources for installation and configuration, including instructions hosted on a GitHub repository (github.com/inference-sh/skills).
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 7, 2026, 10:13 AM
Security Audit — agent-trust-hub — python-executor