Skills
skills/parcadei/continuous-claude-v3/second-order-odes/Gen Agent Trust Hub

second-order-odes

Pass

Audited by Gen Agent Trust Hub on Apr 5, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the Bash tool to execute Python code via uv run. It calls a local script scripts/sympy_compute.py and runs inline snippets using scipy.integrate. These commands are used to solve mathematical problems based on user input.- [PROMPT_INJECTION]: The skill is subject to indirect prompt injection. User-provided mathematical expressions are interpolated into shell commands (e.g., scripts/sympy_compute.py solve "[USER_INPUT]"). There are no boundary markers or sanitization steps defined to prevent a malicious user from escaping the string quotes to execute arbitrary shell commands.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 5, 2026, 01:57 PM