Skills
Navigate
skills/midudev/autoskills/python-executor/Gen Agent Trust Hub

python-executor

Pass

Audited by Gen Agent Trust Hub on Nov 8, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill facilitates the execution of arbitrary Python code on remote infrastructure provided by inference.sh. The execution is managed through the belt CLI utility.
  • [EXTERNAL_DOWNLOADS]: The skill instructs the user to install the belt-sh/cli via npx and references setup documentation and scripts from the inference-sh/skills repository on GitHub.
  • [COMMAND_EXECUTION]: Shell commands are used via the belt tool to authenticate, run remote applications, and pass code payloads for execution.
  • [INDIRECT_PROMPT_INJECTION]: The skill includes a vulnerability surface for indirect prompt injection because it accepts raw Python code in the code input field. If an agent processes untrusted external data and interpolates it into this field without proper sanitization or boundary markers, the remote environment could execute unintended logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 8, 2026, 10:44 PM