The Agent Skills Directory

[COMMAND_EXECUTION]: The script checker/check.py executes git commands via subprocess.run to verify if project files are correctly tracked in version control. These calls are necessary for the tool's conformance checking logic and are implemented using argument lists to prevent shell injection vulnerabilities.- [DYNAMIC_EXECUTION]: The checker processes project YAML files, such as compose.yaml, using yaml.load with yaml.SafeLoader. This is an appropriate and secure method for inspecting configuration files without the risk of executing arbitrary code embedded in the data.- [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection because it processes untrusted repository content provided by the user. 1. Ingestion points: checker/check.py reads and analyzes files and directory structures within a target project path. 2. Boundary markers: The tool does not use explicit delimiters or safety instructions when processing file content. 3. Capability inventory: The tool performs local file reads, executes git commands via subprocess, and parses YAML configurations. 4. Sanitization: The use of SafeLoader for YAML and list-based subprocess arguments mitigates the risk of direct command or code injection.

typo3-site-conformance