The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill facilitates the creation and runtime execution of Python scripts using Frappe's safe_exec environment. These scripts are used to define custom validation rules and are triggered automatically during document lifecycle events such as on_change and before_submit. Examples in references/script-examples.md demonstrate the use of dynamic logic for thresholds, item checks, and multi-level approvals.
[INDIRECT_PROMPT_INJECTION]: The skill processes external, potentially untrusted document data through the doc variable injected into the script context. \n
Ingestion points: The doc object passed to evaluate_document_reviews in references/document-review-system.md. \n
Boundary markers: Absent. \n
Capability inventory: Scripts have access to the Frappe API, including frappe.db queries and safe_exec for logic execution. \n
Sanitization: Absent; no specific sanitization of document input data is enforced before processing within custom scripts.
[COMMAND_EXECUTION]: The documentation provides instructions for executing administrative commands via the bench CLI and performing direct database queries using frappe.db.sql within scripts, as seen in references/troubleshooting.md and references/script-examples.md.

frappe-tweaks-document-review-expert