The Agent Skills Directory

[SAFE]: The skill is a well-structured developer utility that follows Frappe best practices for standalone scripts. It restricts its actions to the local filesystem and provides transparency about the data it extracts.
[COMMAND_EXECUTION]: The agent is instructed to run local Python snippets and shell commands to facilitate database queries and the extraction process. These operations are scoped to the local environment and are necessary for the skill's utility.
[PROMPT_INJECTION]: The skill identifies an indirect prompt injection surface by mirroring database content to the local filesystem.
Ingestion points: Data enters the context in scripts/extract.py via frappe.get_all calls to the database.
Boundary markers: The skill does not use boundary markers or explicit "ignore instructions" warnings when writing extracted code to the filesystem.
Capability inventory: The skill performs filesystem writes (Path.write_text) and directory management (shutil.rmtree, shutil.copy2) within scripts/extract.py.
Sanitization: Path components are sanitized using regex-based slugification in the slugify function to prevent path traversal, though record content is extracted as-is.

frappe-live-code-extractor