What does this skill do, and when should you use it?
This skill focuses on files, images, PDFs, attachments, uploads, downloads, and multimodal inputs in n8n. It separates structured data in $json from file bytes in $binary, then explains how to read, write, and preserve binary data across transformations. For AI Agent tools, it uses storage keys or URLs because tool arguments and returns are JSON-only; for chat surfaces, it uses fetchable HTTPS image URLs. It is a strong fit for n8n and n8n-mcp users who regularly debug missing attachments or unusable file inputs.
Explains how HTTP Request, disk, cloud-storage, email, and AI media nodes populate $binary; shows how to use binaryPropertyName, responseFormat, and getBinaryDataBuffer; demonstrates reattaching binary data in Code node output; preserves files with pass-through settings or Merge in combineByPosition mode; defines inbound and outbound file-staging patterns for AI Agent tools; explains how to upload generated images to object storage or drive-style storage and return their URLs to chat surfaces; and recommends n8n_test_workflow plus n8n_executions to verify that binary data survived execution.
- An n8n developer downloads a PDF or image and finds that its contents were placed in $json instead of $binary.
- An automation engineer uses Code, IF, or Edit Fields and later discovers that an email or upload node has no attachment.
- An AI Agent builder needs a tool to read a user-uploaded file or return a generated file to the user.
- A chatbot developer needs generated images to render in Slack, Discord, Teams, Telegram, or embedded Webhook Chat.
- An n8n troubleshooter needs to identify the node where a binary slot was silently removed.
How do you install this skill?
- Do not assume the documented field names and Merge configuration apply unchanged to every n8n version; confirm them on the target version and run the workflow.
- Use private, least-privilege, short-lived storage for uploaded files and verify that the selected external storage is reachable in the target network environment.
- Treat the skill as guidance and examples, not a substitute for reviewing credential scope, URL exposure, retention, error branches, and cleanup.
- MCP Server
n8nn8n-mcp MCP server
This skill is part of the 15-skill czlonkowski/n8n-skills collection. For Claude Code plugin installation, run: /plugin install czlonkowski/n8n-skills. For manual installation, clone the repository with git clone https://github.com/czlonkowski/n8n-skills.git, then copy the target folder with cp -r n8n-skills/skills/n8n-binary-and-data ~/.claude/skills/. The README also describes downloading and uploading an individual skill folder in Claude.ai; detailed API/SDK installation steps are not included in the supplied material.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/czlonkowski/n8n-skills.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/skills/n8n-binary-and-data" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Once installed, send your agent any of these to trigger it:
- How do I read a PDF in n8n and make it available to an AI Agent tool?
- Why did my attachment disappear after Merge?
After installation, ask a file-related n8n question such as: “How do I read a PDF in n8n and make it available to an AI Agent tool?” or “Why did my attachment disappear after Merge?” The skill activates for files, images, PDFs, attachments, $binary, binaryPropertyName, PDF reading, image sending, Merge-related binary loss, and CDN requirements for chat images. With n8n-mcp, use get_node to confirm node fields and n8n_test_workflow plus n8n_executions to inspect actual execution output.
What are this skill's strengths and limitations?
- Clearly explains the $json versus $binary split and the most common causes of file loss.
- Provides concrete patterns for HTTP downloads, Code nodes, Merge, AI Agent tools, and chat images.
- Explicitly documents the JSON-only AI tool boundary and the URL requirement for chat clients.
- Includes anti-patterns, verification steps, and integration points with related n8n skills.
- Assumes an n8n workflow environment and is designed to work with n8n-mcp tooling.
- Does not provide a CDN, object store, or file host; users must supply and configure storage.
- Full Data Tables coverage, Code node sandbox details, and Custom Code Tool details belong to other skills.
- The supplied material provides no test results for this individual skill across n8n versions or standalone platforms.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| n8n Binary & File Handling this page | 53 · Use with care | ★ 6.4k | 19d ago | MIT |
| n8n Error Handling | 53 · Use with care | ★ 6.4k | 19d ago | MIT |
| n8n AI Agent Architecture Guide | 50 · Use with care | ★ 6.4k | 19d ago | MIT |
| n8n Subworkflow Architect | 49 · Use with care | ★ 6.4k | 19d ago | MIT |
| n8n Workflow Architecture Patterns | 40 · Not recommended | ★ 6.4k | 19d ago | MIT |
How did FollowSkills review this skill?
The skill clearly separates $json from $binary, recommends private storage, signed short-lived URLs, avoiding hardcoded filenames, and surfacing upload and 404 failures. Deduction is warranted because least-privilege permissions, credential isolation, explicit user confirmation, rollback, and deletion recovery are not fully specified; external storage and public URLs can still expose data.
The core concepts, code examples, Merge synchronization pattern, executeOnce guidance, and anti-patterns are broadly consistent, and the skill tells users to inspect executions. Deduction is required because this is static-only evidence with no runnable tests or CI proof; many fields are version-dependent and require separate get_node confirmation, while error handling is referenced more than fully specified.
Triggers cover files, PDFs, images, uploads, downloads, $binary, AI tools, and chat image rendering. The skill also distinguishes agent and non-agent flows and discloses platform-specific seams. Deduction reflects incomplete systematic non-fit boundaries, no explicit Chinese-language support, and no evidence of reachability for external storage services from mainland China.
The material is layered across the main skill, focused reference files, README, checklists, examples, anti-patterns, integrations, limitations, MIT licensing, and version 1.0.0. Deduction reflects missing maintenance ownership, changelog, and explicit update path; the repository context has inconsistent skill-count descriptions, and important behavior depends on referenced files outside the selected path.
The skill supplies directly reusable n8n JSON, JavaScript, Merge, and storage-key patterns for binary retention, agent-tool transfer, and chat rendering. Deduction reflects the absence of executed representative outputs; users still need version checks, credential and storage configuration, platform-specific mapping, and runtime verification.
The README and skill identify n8n documentation as authoritative sources and prescribe n8n_test_workflow plus n8n_executions for checking binary survival. Deduction is necessary because the supplied evidence is static only, with no verifiable CI, committed test suite, or independent reproduction record; key claims have limited corroboration.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →