The Agent Skills Directory

[PROMPT_INJECTION]: The skill's instructions and reference documentation do not contain any patterns intended to bypass safety guardrails or override agent behavior. The instructional language is strictly professional and focused on incident management procedures.
[DATA_EXFILTRATION]: No network-capable code or attempts to access sensitive system files (e.g., SSH keys, AWS credentials) were found. The Python scripts operate locally on provided data and do not use network libraries like requests or urllib.
[REMOTE_CODE_EXECUTION]: The skill does not download or execute code from external sources. All scripts are self-contained and use only the Python standard library, as stated in the project quality standards.
[COMMAND_EXECUTION]: There is no evidence of arbitrary command execution. The scripts do not utilize functions like os.system, subprocess.run, or eval to interact with the shell.
[CREDENTIALS_UNSAFE]: The codebase was scanned for hardcoded API keys, tokens, and secrets. None were found. Placeholders in communication templates (e.g., {api_key}) are used appropriately for illustrative purposes.
[SAFE]: The skill is a utility for data processing and documentation. It follows the principle of least privilege, requiring only file or stdin access to perform its functions. The inclusion of comprehensive reference documentation and sample data supports its intended use case without security risk.

incident-commander