The Agent Skills Directory

[SAFE]: The skill provides a robust framework for PHP modernization using industry-standard tools and practices. All components are aligned with their stated purpose of improving code quality and version compatibility.
[COMMAND_EXECUTION]: The included Python scripts (introspect.py, modernize_loop.py, verify_php_project.py) execute local PHP development binaries discovered in the project's vendor/bin or .Build/bin directories. These executions are performed using subprocess.run with list-based arguments, avoiding shell injection risks and restricted to the project root.
[EXTERNAL_DOWNLOADS]: The skill references official PHP documentation and utilizes well-known GitHub Actions from trusted organizations like Astral and GitHub for its CI/CD templates. These are recognized as safe external resources under the established trust scope.
[INDIRECT_PROMPT_INJECTION]: The skill analyzes external PHP project files, which constitutes a legitimate ingestion surface for its intended function.
Ingestion points: Project configuration files (e.g., composer.json) and PHP source code.
Boundary markers: The scripts produce structured JSON and SARIF reports to isolate analysis results.
Capability inventory: Subprocess execution of discovered PHP binaries and file system operations.
Sanitization: The skill relies on the underlying tools' validation, and the agent is instructed to review all proposed changes before they are applied.

php-modernization