The Agent Skills Directory

[SAFE]: The skill content consists of standard instructional material and best practices for shell scripting. It does not contain prompt injections, obfuscation, or malicious code.- [INDIRECT_PROMPT_INJECTION]: The skill is intended to process and refactor user-provided shell scripts, creating a potential attack surface. However, the instructions explicitly include a 'Security Patterns' section that directs the agent to implement input validation, avoid eval, and use safe temporary file patterns, which mitigates these risks in the output code.- [COMMAND_EXECUTION]: The allowed-tools configuration includes shellcheck and bash -n, which are legitimate static analysis tools used for linting and syntax checking of shell scripts. These tools are restricted to their intended purpose and do not pose a risk in this context.

google-shell-style