Civil Engineering Project Planning, Now Just Tell Your AI CPM (Construction Project Master) — A free MCP server connecting Claude & ChatGPT to full project planning automation

-
🏗 22B Labs · Open Source Release

Civil Engineering Project Planning, Now Just Tell Your AI

CPM (Construction Project Master) — A free MCP server connecting Claude & ChatGPT to full project planning automation

📅 April 2026 ✍️ 22B Labs ⏱ 5 min read

Every civil engineering project starts the same way. You call a consultant for a rough cost estimate. You dig through ministry guidelines to confirm permitting requirements. You build the BOQ spreadsheet by hand. Again. Engineers at small municipalities or contractors do it alone.

CPM changes that to a single conversation. Tell Claude or ChatGPT what you want to build — it handles cost estimation, legal procedures, BOQ Excel, investment report DOCX, and drawings automatically.

"The goal of this project is to reduce the inequality of access to expert knowledge.
Anyone should be able to run a proper preliminary project review — no consulting budget required."
system overview

How It's Built

CPM system overview — AI client to MCP server architecture
① System Overview — Claude / ChatGPT connects to CPM MCP server and calls 19 tools directly.
workflow

Automated Workflow

CPM workflow — from natural language input to output generation
② Full Workflow — Natural language → parse → quantity estimation → unit pricing → document generation, all automatic.
Project phase diagram
③ Phase Diagram — Planning → Permits → Design → Procurement → Construction → Completion.
Process flow diagram
④ Process Flow — Tool call sequence and data flow.
output

What Actually Comes Out

Everything below was automatically generated from: "Plan a new local road L=890m, B=6m, Gyeonggi Province, hilly terrain."

Project cost summary — breakdown by trade
⑤ Cost Summary — 8 trade costs + design/supervision/contingency fees. Total project cost auto-calculated.
Project timeline
⑥ Project Timeline — Permits, design, procurement, construction, completion phases auto-scheduled.
Comprehensive view of all outputs
⑦ Comprehensive View — Cost, quantities, procedures, and schedule in one unified view.
Output examples — BOQ Excel, investment report DOCX, SVG drawings
⑧ Output Examples — BOQ Excel · Investment Report DOCX · SVG Drawings, all auto-generated.
Document generation — Word investment report
⑨ Document Generation — Investment report DOCX with cover, overview, costs, legal procedures, and expected effects.
Detailed workflow — per-tool inputs and outputs
⑩ Detailed Workflow — 19-tool call sequence with inputs and outputs.
Auto-generated plan view — 890m road alignment
⑪ Plan View — 890m road alignment SVG auto-generated. Stations, intersection points, cut/fill zones, utility pipe routes placed automatically.
features

19 MCP Tools

💰
Preliminary Cost Estimate
Korean standard specs, PPS market prices, wage rates. ±20–30% accuracy.
📋
Legal Procedure Guidance
16 ministry guidelines built-in. Investment review, permits, EIA auto-identified.
📊
BOQ + Investment Report
Auto-generate Excel BOQ and DOCX investment report.
🗺️
Land Info Query
Address → zoning / BCR / FAR / land price auto-query.
📐
Drawing Output
Cross-section + plan view SVG. DXF (AutoCAD) in v1.1.
📈
Feasibility Analysis
IRR, NPV, DSCR, sensitivity analysis. (v1.1)
comparison

Why Existing Approaches Fall Short

ApproachTimeCostLimitation
Consulting firm1–4 weeksMillions of KRWOnly for well-funded organizations
Manual preparationDays to weeksStaff hoursRequires deep expertise, error-prone
Estimating softwareHoursPaid subscriptionDetailed design phase only
CPM (MCP)MinutesFreeNatural language, AI handles everything
how to use

How It Works

# In Claude Desktop or ChatGPT: "Plan a new local road, L=890m, B=6m, Gyeonggi Province, hilly terrain, 2026–2028." → Preliminary cost estimate → BOQ Excel auto-generated → Investment report DOCX auto-generated → Plan view + cross-section drawings → Legal procedure checklist
💡 Connect to Claude Desktop — Add to claude_desktop_config.json and restart:

{ "mcpServers": { "cpm": { "command": "mcp-remote", "args": ["http://127.0.0.1:8765/mcp"] } } }
⚠️ Disclaimer
All CPM outputs are preliminary estimates (±20–30%) for internal review only. Cannot be used for official submissions, bidding, or contracts.

Get Started on GitHub

MIT License · Completely Free · 19 MCP Tools · Python
Works with Claude Desktop + ChatGPT

🔗 View on GitHub
#CPM #MCP #CivilEngineering #ProjectPlanning #OpenSource #Claude #ChatGPT #22BLabs #The4thPath
22
22B Labs · The 4th Path GitHub: sinmb79 · the4thpath.com
The 4th Path: ⟨H⊕A⟩ ↦ Ω — Human × AI → a better world.

Comments

Post a Comment

Popular Now

Paperclip AI Review: "If Agents Are Employees, This Is the Company"

-
Open Source Review · 2026 Paperclip AI Review: "If Agents Are Employees, This Is the Company" An open-source orchestration platform for building zero-human companies. 32K GitHub stars in weeks, a marketplace on the way, and a concept that's redefining what it means to run a business. Here's the full picture — strengths, gaps, market reception, and where this is all going. Reviewed March 25, 2026 Author 22B Labs · The 4th Path Read time ~10 min Repo paperclipai/paperclip Overall Score 8.7 ★★★★☆ Verdict: Highly Recommended for Solo Founders & Multi-Agent Builders The most complete open-source answer to "how do I manage 20 AI agents at once." Early-stage rough edges, but the architecture is sound and momentum is undeniable. Contents what it is core features market reception use cases vs competitors roadmap risks & limits final verdict What Is Paperclip? Paperclip is an open-source Node.js server with a R...
Read more

oh-my-openagent (OmO) — Full Review: "The Multi-Model Harness That Escaped Claude's Prison"

-
Open Source Review · 2026 oh-my-openagent (OmO) — Full Review: "The Multi-Model Harness That Escaped Claude's Prison" An OpenCode plugin that transforms a single AI coding assistant into a virtual dev team of 11 specialized agents working in parallel. 37.6K GitHub stars, a history of being blocked by Anthropic, and an open-source rebellion built by a Korean developer. Full breakdown — from install to real-world use. Published March 25, 2026 Author 22B Labs · The 4th Path Read time ~12 min Repo code-yeongyu/oh-my-openagent Overall Score 8.4 ★★★★☆ Best for: Multi-model explorers, heavy terminal coders, cost-conscious builders Genuinely more powerful than vanilla Claude Code for complex multi-file work. The prerequisite: you understand the cost implications and know when not to use it. A project with strong personality and real practical value. Contents what is OmO origin story 11 agents how to use use cases vs competitors what...
Read more

GitHub's VS Code BYOK move matters more than it looks

-
Bring-your-own-key support can look like a small settings change until you view it through procurement, governance, and model volatility. GitHub's new BYOK option for Copilot Business and Enterprise users in VS Code signals a larger shift in developer tooling. Teams increasingly want the convenience of an integrated AI surface without surrendering model choice, billing control, or policy flexibility to a single vendor path. Three Things to Know GitHub now lets organizations use their own model-provider keys inside VS Code Chat and agent workflows, including built-in and custom agents. That matters because enterprise teams want portability across providers, cleaner cost accounting, and the option to use local or regional models. Viewed alongside GitHub's recent interaction-data policy update, BYOK also reads as a trust feature for organizations that care deeply about data boundaries. A small product change with a real governance message GitHub's April 22 changelog entry...
Read more