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
- ActiveCampaign/activecampaign-plugintheir own version, first seen Oct 20, 2026
- opensite-ai/opensite-skillstheir own version, first seen Oct 13, 2026
- aiunlocked1412/claude-skill-unlocktheir own version, first seen Apr 16, 2026
- accolver/skill-makertheir own version, first seen Mar 10, 2026
Other versions4 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 4
- aiunlocked1412/claude-skill-unlock★ 18same name
- opensite-ai/opensite-skills★ 15same name
- accolver/skill-maker★ 4same name
- ActiveCampaign/activecampaign-plugin★ 4same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository10 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- business-researchSkill
- competition-demoSkill
- execution-planSkill
- lordgen-3d-web-experienceSkill
- lordgen-full-rebuildSkill
- lordgen-site-upgradeSkill
- opportunity-scoreSkill
- proposal-generatorSkill
- workflow-qaSkill
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
- Very specific: line 15
Read data/execution-plan.json in full, especially open_dependencies — do not build around a dependency that's still unresolved (e.g.
- Very specific: line 34
Every node the plan calls for as an error-handling or validation step must actually exist in the built workflow — cross-check against …
“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.
Files
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)
- 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'sexecution-plan.md's job, already done. - Read
data/execution-plan.jsonin full, especiallyopen_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:
search_nodes— find the actual nodes needed (trigger type, HTTP/webhook, AI/LLM node, notification node) rather than guessing node names.get_workflow_best_practices— pull n8n's own guidance for the relevant technique (e.g. webhook triggers, error handling) before writing SDK code.- Draft the workflow as n8n Workflow SDK code, matching
data/execution-plan.json'sworkflow_stepsone-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). validate_workflow(andvalidate_node_configfor individual nodes) — fix errors before creating anything in n8n.create_workflow_from_code— create the workflow in n8n, left unpublished/deactivated.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.- Export/describe the result into
workflows/competition-demo.jsonand report it for human review.
Hard rules
- Never call
publish_workfloworexecute_workflowagainst real external services without the developer's explicit, in-the-moment approval.test_workflowwith pin data is the only execution mode this skill may use autonomously — it's specifically designed not to touch live services. This mirrorsCLAUDE.md§10 and the Human Approval Point already written intoexecution-plan.md. - Never silently substitute a different node/service than what
execution-plan.mdspecifies. 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_draftonly, 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_handlingbefore 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.
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
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.

