The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/with_server.py uses subprocess.Popen with the shell=True argument to execute strings passed via the --server command-line option. This pattern allows for the execution of arbitrary shell commands and increases the risk of command injection.\n- [PROMPT_INJECTION]: The SKILL.md file contains a directive ('DO NOT read the source until you try running the script first') that instructs the AI agent to ignore the script's implementation details. This is a self-referential instruction designed to influence the agent's behavior by discouraging the inspection of code before it is executed, which undermines the agent's internal safety reasoning and security protocols.

webapp-testing