The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local Python scripts (validate_review_report.py, render_review_report_html.py, and validate_code_comment_directives.py) using uv run. These scripts are used for structural validation and rendering of the final review artifacts. The commands are scoped to the skill's own internal script directory.
[DATA_EXFILTRATION]: The skill reads from and writes to the local file system to manage review artifacts (e.g., /tmp/wibaek-review/ and <repo_root>/.codex/reviews/). These operations are consistent with the skill's stated purpose of generating and storing engineering reports. No network operations or external data transfers were identified.
[PROMPT_INJECTION]: The instructions provide a structured workflow for data assembly and report generation. There are no patterns suggesting attempts to override model constraints, bypass safety filters, or extract system prompts.
[REMOTE_CODE_EXECUTION]: All executable code is contained within the skill's own script directory. There are no patterns of downloading external scripts via curl or wget for immediate execution.

wibaek-review-final-report