frappe-core-files

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt includes examples that embed API keys and secrets directly (e.g., curl Authorization header and site_config.json with s3_access_key/s3_secret_key), which would require the model to insert secret values verbatim into commands or configs.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.65). The required workflow includes save_url(file_url, ...) / “Save File from URL”, which at runtime fetches external URL content (outsider-authored web text/binary converted to readable content) and then stores/returns it via the File API, making it available to the agent’s LLM context through subsequent frappe.get_file(...)/file_doc.get_content()-style reads.