The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified in the provided documentation and code examples. The skill is purely educational and focuses on correct framework usage.
[PROMPT_INJECTION]: No instructions were found that attempt to override agent behavior, bypass safety filters, or extract system prompts.
[DATA_EXFILTRATION]: No network operations or patterns involving the exfiltration of sensitive data were detected. The skill uses standard internal logging mechanisms (frappe.log_error).
[COMMAND_EXECUTION]: The skill does not contain any shell commands, subprocess calls, or patterns that would lead to arbitrary command execution.
[CREDENTIALS_UNSAFE]: No hardcoded credentials, API keys, or sensitive environment variable access was found. The examples use placeholders where appropriate.
[REMOTE_CODE_EXECUTION]: The skill does not perform any external downloads or execute code from remote sources.
[INDIRECT_PROMPT_INJECTION]: While the skill describes how to process document data within permission logic (an ingestion surface), it provides explicit security guidance on using frappe.db.escape() to prevent SQL injection and using frappe.get_list() to respect permission boundaries. It does not expose dangerous capabilities to untrusted input.

frappe-errors-permissions