Skills
skills/starchild-ai-agent/official-skills/script-generator/Gen Agent Trust Hub

script-generator

Fail

Audited by Gen Agent Trust Hub on Mar 14, 2026

Risk Level: HIGHCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [COMMAND_EXECUTION]: The skill directs the agent to execute the Python scripts it generates by using the bash() function (e.g., bash("python3 scripts/<name>.py")). This allows for the execution of arbitrary shell commands.
  • [REMOTE_CODE_EXECUTION]: The core functionality of the skill is to write Python logic to local files and then execute them immediately or schedule them for later. This allows a user to indirectly execute arbitrary code on the host system via the agent.
  • [DATA_EXFILTRATION]: The documentation explicitly lists sensitive environment variables (TAAPI_API_KEY, COINGLASS_API_KEY, LUNARCRUSH_API_KEY, COINGECKO_API_KEY) and instructs the scripts to access them using os.getenv(). Because scripts use the requests library, an attacker could easily transmit these credentials to an external server.
  • [COMMAND_EXECUTION]: The schedule_task function is used to create recurring background processes (python3 scripts/foo.py). This capability allows for persistent execution of code on the system without further agent intervention.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Mar 14, 2026, 10:07 PM