The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill documents the use of n8n nodes to download files from remote servers and cloud storage providers such as AWS S3, Google Drive, and Cloudflare R2. These operations are native to the platform and are described in a purely instructional context.\n- [COMMAND_EXECUTION]: The skill provides code templates for n8n's Code node, showing how to use JavaScript to interact with file buffers via platform-provided helpers. It also covers the configuration of network-capable nodes for data transfer.\n- [DYNAMIC_EXECUTION]: The skill describes how to programmatically generate file metadata and encode file contents in Base64 at runtime within n8n scripts, which is the standard procedure for file handling in the platform.\n- [INDIRECT_PROMPT_INJECTION]: The skill identifies the attack surface presented by external file ingestion points like chat triggers, email attachments, and webhooks. It provides a detailed security architecture to mitigate these risks, including: (1) Ingestion points: Trigger files and HTTP downloads; (2) Boundary markers: Clear guidance on the agent-tool JSON boundary; (3) Capability inventory: Node-based file and network operations; (4) Sanitization: Recommending private storage, signed URLs, and session-scoped keys to prevent unauthorized access and path traversal.

n8n-binary-and-data