agentleFS
Kikobazz123Kikobazz123/lordgen-competition-demo/.claude/skills/automation-builder/SKILL.md

automation-builder

Skill0 starsChanged 3 months agoMITView on GitHub

Translate data/execution-plan.json into a real, tested n8n workflow using n8n's MCP server, without ever activating or executing it against live services without explicit human approval. Use after execution-plan has produced data/execution-plan.json AND n8n MCP is connected.

4 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

2 rules. 0 are common, 0 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 4 edited versions we read keep these.

  • 0

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read data/execution-plan.json in full, especially open_dependencies — do not build around a dependency that's still unresolved (e.g.

    line 15
  2. Very specific:

    Every node the plan calls for as an error-handling or validation step must actually exist in the built workflow — cross-check against …

    line 34

“Common” is measured: kept by at least 6 in 10 of the 4 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Automation Builder

Purpose

Turn the execution plan into a real starter automation in n8n — not a hand-written JSON guess at what n8n workflows look like. This skill only works once n8n's MCP server is connected (docs/architecture.md); it must not fabricate n8n tool calls or workflow structure if the server isn't available.

Precondition check (do this first, every time)

  1. Confirm n8n MCP tools are actually available (e.g. via ToolSearch). If not connected, stop and say so — do not simulate what the workflow "would" look like as if it were built; that's execution-plan.md's job, already done.
  2. Read data/execution-plan.json in full, especially open_dependencies — do not build around a dependency that's still unresolved (e.g. an unidentified source system) without flagging it.

Method (n8n's own documented MCP tools — see docs/n8n-reference.md for the full list and source)

Build in this order, using n8n's Workflow Builder tools:

  1. search_nodes — find the actual nodes needed (trigger type, HTTP/webhook, AI/LLM node, notification node) rather than guessing node names.
  2. get_workflow_best_practices — pull n8n's own guidance for the relevant technique (e.g. webhook triggers, error handling) before writing SDK code.
  3. Draft the workflow as n8n Workflow SDK code, matching data/execution-plan.json's workflow_steps one-to-one — every step in the plan should be a traceable node or node group, and vice versa (no extra nodes the plan didn't ask for, no missing steps).
  4. validate_workflow (and validate_node_config for individual nodes) — fix errors before creating anything in n8n.
  5. create_workflow_from_code — create the workflow in n8n, left unpublished/deactivated.
  6. prepare_workflow_pin_data + test_workflow — test using pin data, which bypasses real external services (per n8n's own docs). This is how "Testing" in the execution plan actually gets exercised, not simulated.
  7. Export/describe the result into workflows/competition-demo.json and report it for human review.

Hard rules

  • Never call publish_workflow or execute_workflow against real external services without the developer's explicit, in-the-moment approval. test_workflow with pin data is the only execution mode this skill may use autonomously — it's specifically designed not to touch live services. This mirrors CLAUDE.md §10 and the Human Approval Point already written into execution-plan.md.
  • Never silently substitute a different node/service than what execution-plan.md specifies. If the right node doesn't exist or the plan's approach isn't buildable as described, stop and report the discrepancy (CLAUDE.md §8) rather than quietly building something adjacent.
  • The Gmail send step stays a draft, not a send, consistent with the Gmail MCP connector available in this environment (create_draft only, no send tool) — the workflow should produce a draft for a human to send, or clearly mark where a real send capability would need to be added and approved.
  • Every node the plan calls for as an error-handling or validation step must actually exist in the built workflow — cross-check against data/execution-plan.json.error_handling before calling this skill done.

Required Output

  • The workflow created in n8n (unpublished)
  • workflows/competition-demo.json — exported/described workflow, for the repo and for Phase 8 (Integrations/handoff)
  • A short build report: which execution-plan steps map to which nodes, what was tested (with pin data) and the result, and any plan items that couldn't be built as specified

Out of Scope

This skill does not connect ClickUp or decide human-approval UX beyond what's already in execution-plan.md — those are Phase 7 (Integrations). It does not activate the workflow for production use under any circumstance.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.