The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local Python script (scripts/generate_impact_decision.py) which in turn invokes the git command-line tool via subprocess.run. This is used to perform git apply --check to determine if security patches are compatible with the current codebase. The command is constructed using a list of arguments and does not invoke a shell, which mitigates standard command injection risks. This behavior is integral to the skill's stated purpose of impact analysis.
[SAFE]: The skill operates entirely on local data found in the .ace-outputs directory and the project's source tree. It does not perform any network operations, access sensitive credentials, or attempt to persist across sessions.
[SAFE]: Data ingestion is limited to structured JSON and Markdown files produced by previous steps in the security analysis pipeline. The processing logic in the Python script is based on predefined regex patterns and file path verification, which reduces the risk of indirect prompt injection affecting the analysis outcome.

arkweb-security-patch-impact