The Agent Skills Directory

[COMMAND_EXECUTION]: The skill establishes a workflow for executing verification tests and recording their output. To mitigate risks, it explicitly instructs the agent to consult a configuration file (AGENTS.md) for defined commands rather than improvising or guessing based on external input.
[PROMPT_INJECTION]: The instructions do not contain attempts to bypass safety filters or override system instructions. Instead, they define operational constraints to ensure the agent stays within the intended scope of assigned tasks.
[SAFE]: The skill promotes security and reliability best practices, including maintaining isolated work environments (separate branches or worktrees) and requiring adversarial self-review of all code changes before completion.

implement-task