The n8n Nodes Catalog: a complete, machine-readable reference

A free, structured dataset of every n8n node, parsed from source and refreshed monthly. Built for LLM training, AI agent grounding, and developer reference.

Fix UiPath Studio CS0246 "GlobalVariablesNamespace" publish error

CS0246 GlobalVariablesNamespace fires at Publish even when Studio runs the project clean. The fix has four paths: re-add the namespace, prefix references, restore the .local folder, or move the project off a non-NTFS path. Match your variant from the table.

agency-os: the Notion board that plans your work, then ships it

We turned Notion into the dashboard of a small AI agency. You drop ideas in. Agents plan the work, sequence it, and ship the output. Here's why we built it and how it works.

How to fix CrewAI hierarchical DelegateWorkToolSchema validation error

Upgrade CrewAI to a release containing PR #2608. The schema accepts Union[str, dict] and _run coerces dict args back to strings. Four failures look the same, four fixes.

n8n MCP server: build, lint, and debug workflows from your AI agent

Nine tools that generate correct n8n JSON, lint for the failure modes that matter, and diagnose silent data loss. Install snippets for Cursor, Claude Code, and six more hosts.

How to fix LangChain MultiServerMCPClient NotImplementedError on Windows

MultiServerMCPClient raises NotImplementedError on Windows because asyncio does not support Unix sockets. Fix: pass transport="streamable_http" or run inside WSL2. Full workaround inside.

How to fix Claude Code SessionStart hooks firing without context

Plugin SessionStart hooks fire but Claude does not see additionalContext. Move the hook to ~/.claude/settings.json. Plus four similar bugs that need different fixes.

n8n Code node: Run Once for All Items vs Each Item

Run Once for All Items uses $input.all() and returns an array; Run Once for Each Item uses $json and returns { json: ... }. Three traps to watch.

How to fix the Make HTTP module 40-second timeout error

Make's HTTP module times out individual requests at roughly 40 seconds. The fix is two-layered: retry transient slowness with Break, and redesign the call when the API is reliably slow.

What is the Model Context Protocol (MCP)? A 2026 primer

MCP is the on-demand counterpart to event-driven automation. Zapier triggers workflows; MCP gives the AI hands to call them when asked.