Borrowing it
Nothing to install: this file belongs to neurawork-git/n8n-autopilot. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/neurawork-git/n8n-autopilot/main/CLAUDE.mdgit clone --depth 1 https://github.com/neurawork-git/n8n-autopilotWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/neurawork-git/n8n-autopilot/claude-md)<a href="https://agentmods.dev/instructions/neurawork-git/n8n-autopilot/claude-md"><img src="https://agentmods.dev/badge/instructions/neurawork-git/n8n-autopilot/claude-md/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/instructions/neurawork-git/n8n-autopilot/claude-md"><img src="https://agentmods.dev/badge/instructions/neurawork-git/n8n-autopilot/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.03312 | $0.03312 |
| Opus 5 | $0.01656 | $0.01656 |
| Sonnet 5 | $0.00662 | $0.00662 |
| Haiku 4.5 | $0.00331 | $0.00331 |
Grade A, and why
n8n-autopilot CLAUDE.md scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- **DataTable CRUD** — no `datatable` subcommand; use the `/n8n-autopilot:data-tables` skill (curl carve-out for `/api/v1/data-tables` only). How it starts
The opening of the file, as written. The whole thing — 138 lines — stays where its author put it; the contents beside it link to each section on GitHub.
n8n Workflow Development — n8n-autopilot Plugin
This repo uses the n8n-autopilot plugin for Claude Code. Workflows are TypeScript (Decorator format). Never write n8n JSON by hand.
Reference n8nac version: 2.4.0 (minimum 2.3.0), v4-native environment-centric config model. SSOT =
REFERENCE_N8NAC_VERSIONinscripts/setup-check.sh. Setup + bump procedure: docs/rules/setup.md.
Knowledge skills — read BEFORE inventing CLI surface
| Skill | When to read |
|---|---|
n8nac-cheatsheet |
Default first stop. "Which command for X?" — curated table, 60+ common operations across workspace, env, workflows, executions, credentials, recipes, schemas. |
n8nac-reference |
Raw n8nac --help tree, 61 command/subcommand blocks. Source of truth for "does this command exist?" — if not in reference.md, it does not exist. |
n8n-architect (companion plugin) |
Schema-First Research, workflow authoring rules, AI/LangChain patterns, Common Mistakes. Owned by Etienne's n8n-as-code plugin. |
Rule: before running npx n8nac <cmd> --help interactively, grep the cheat-sheet, then the
reference file. Only fall back to live --help if both fail — and treat that as a sign to regenerate
the reference (bash scripts/dump-n8nac-help.sh > skills/n8nac-reference/reference.md).
Cheat-Sheet — User asks X → run Y
Use these BEFORE inventing flags or grepping --help. If the user's intent matches a row, run
the linked skill / command verbatim. If no row matches, ask the user — do not guess.
| User intent | Exact command |
|---|---|
| "find credential " / "which cred is X" / "list Dropbox creds in this project" | /n8n-autopilot:find-credential <pattern> (project-scoped by default; add --type <credType> or --project all) |
| "list n8n projects" / "which projects exist" / "show projects on instance" | /n8n-autopilot:find-project |
| "switch workspace to project X" | verify env target via npx n8nac env list --json, then npx n8nac env update <env> --project-name "<X>" then /n8n-autopilot:check-mcps |
| "show active project / instance binding" | npx n8nac workspace status --json — project + sync folder only. env-blind: its env field reports the GLOBAL active env, NOT the session env. For "which env am I on" use npx n8nac env list --json (never infer the session env from workspace status) |
| "build a workflow that does X" | /n8n-autopilot:build-workflow "<description>" (prose pipeline) |
| "build/edit with hard-enforced gates" (experimental) | /n8n-autopilot:build-workflow-v2 "<description>" — JS-orchestrated, gate-checks as control flow |
| "build a whole multi-workflow STACK" / "orchestrator + sub-workflows from one use case" (experimental) | /n8n-autopilot:build-stack-v2 "<end-to-end use case>" (or extend "<stack>" "<change>") — decompose → handover contracts + mermaid → bottom-up build via build-workflow-v2 |
| "help me plan an n8n workflow" / "I want to automate X but don't know how" / rough idea only | /n8n-autopilot:stack-intake "<one-line idea>" — guided interview → writes a PRP for build-stack-v2 |
| "deploy .workflow.ts" | /n8n-autopilot:deploy <file>.workflow.ts |
| "pull every remote-only workflow / fix mirror drift" | /n8n-autopilot:mirror-sync |
| "which env/instance is this session on" / "pin session to env X" / "why was my env command blocked" | /n8n-autopilot:session-env (explains model + reports session-vs-global). Pin per-session via .claude/settings.json env block N8NAC_ENVIRONMENT=<env> (or export/--env) — never env use (mutates shared global, blocked by clobber-guard). See docs/rules/gates.md. |
| "fix stale credential IDs in workflows" | /n8n-autopilot:sync-credentials --fix-workflows (project-scoped) |
| "regenerate inventory / list of nodes" | /n8n-autopilot:inventory |
| "data table CRUD" (create/seed/list/drop tables in n8n) | /n8n-autopilot:data-tables |
| "give feedback" / "report friction" / "feedback zum plugin" | /n8n-autopilot:feedback (interview); /n8n-autopilot:feedback sync pushes centrally (consent-gated) |
| "pull schemas" / "update node schemas" | /n8n-autopilot:pull-schemas |
| "check setup / MCP / instance health" | /n8n-autopilot:check-mcps |
| "find a workflow by name" | npx n8nac find <query> --json (use --remote for instance-side, default = local+remote) |
| "test schedule/manual/error-trigger workflow" / "non-HTTP test" | /n8n-autopilot:test-manual <workflowId> (resolves UI URL → waits for execution-id → inspects run) |
| "show executions of workflow " | npx n8nac execution list --workflow-id <id> |
| "inspect a specific execution" | npx n8nac execution get <executionId> --include-data |
| "resolve workflow URL for UI" | npx n8nac workflow present <id> --json |
| "pull workflow from instance" | npx n8nac pull <id> |
| "refresh remote state without overwriting local" | npx n8nac fetch <id> |
"create a NEW credential type" / before credential create |
npx n8nac credential schema <type> FIRST — n8n rejects unknown credential fields with HTTP 400 (e.g. httpHeaderAuth requires allowedDomains). Never guess the credential body shape |
| "activate a stack / parent workflow that calls sub-workflows" | publish each sub-workflow (the executeWorkflow/mcpTrigger target) in the n8n UI BEFORE activating the parent — push does not auto-publish; parent activation fails on an unpublished child |
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 10d ago First seen · 138 lines · 3,312 tokens per session scan A 74fcf7263f9a
n8n-autopilot CLAUDE.md is an instructions file published in the GitHub repository neurawork-git/n8n-autopilot (18 stars, last pushed 1mo ago), licensed MIT. It adds 3,312 tokens to every session, about $0.0166 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
deepseek-harness AGENTS.md
AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.