Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for software architects

Service Boundary and Coupling Agent

A ranked, tested set of boundary changes that reduce coupling and let services change separately

DecisionsIT and development
See how it worksGet access
Service Boundary and Coupling Agent: what goes in, what the agent does and what you get

What it does

A change to one service keeps breaking three others, but nobody has a map of why. This agent reads the code repositories, call graphs, shared database schemas and deployment history. It measures coupling: how many services call each other, which share a database table, which form cycles, and which are changed together in the same release. It ranks the worst hotspots and proposes boundary changes, such as moving a table to one owner, adding an interface, or merging two services that always change together. For each proposal it rebuilds the graph on paper, measures the coupling again and checks that no new cycle appears. If a proposal makes things worse, it drops it and tries another. The architect decides which changes go forward. Edge case: a shared table is read by 11 services.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
Yes, continueYes, continueApprovedNoNo 1 STARTS WHEN Architect requests a coupling review 2 USES A TOOL Read call graphs, schemas and release history 3 DOES Measure calls, shared tables, cycles and changesreleased together 4 DOES Rank the worst hotspots 5 DOES Propose a boundary change for each top hotspot 6 USES A TOOL Rebuild the call graph with the change applied andremeasure 7 CHECKS THE RESULT Does the change lower coupling without creating anew cycle? If not: drop that proposal and try another option forthe hotspot. Back to step 5. 8 DOES Estimate effort and risk for each surviving proposal 9 CHECKS THE RESULT Do the combined changes still leave an acceptableload on each service? If not: rework the heaviest service's proposal. Back tostep 5. 10 YOU APPROVE Architect decides which changes to take forward 11 RESULT Coupling report with ranked proposals
Read the steps as a list
  1. Architect requests a coupling review
  2. Read call graphs, schemas and release history
  3. Measure calls, shared tables, cycles and changes released together
  4. Rank the worst hotspots
  5. Propose a boundary change for each top hotspot
  6. Rebuild the call graph with the change applied and remeasure
  7. Does the change lower coupling without creating a new cycle?If not: drop that proposal and try another option for the hotspot. Back to step 5.
  8. Estimate effort and risk for each surviving proposal
  9. Do the combined changes still leave an acceptable load on each service?If not: rework the heaviest service's proposal. Back to step 5.
  10. Architect decides which changes to take forwardThe agent waits here for your OK.
  11. Coupling report with ranked proposals

How it decides

It prefers proposals that cut cross-service calls and shared tables without creating a new cycle or a heavier load on one service.

  • Rank shared tables read by more than 3 services as top risk
  • Treat two services changed together in over 70 percent of releases as merge candidates
  • Reject a proposal that creates a call cycle
  • Show effort next to every gain

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Repositories to scan
  • Coupling thresholds
  • Release history window (default 12 months)
  • Output format

What keeps you in control

It always asks you first

  • Which proposals go forward

Hard limits

  • Never edits code or schemas
  • Never presents a guess as measured data

It stops when

  • Done: ranked proposals with tested effects delivered
  • Stop: call data is missing for key services

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensThe agent found the orders table was read by 11 services and Billing and Invoicing changed together in 82 percent of releases. It proposed giving Orders one owner with a read interface, and merging Billing and Invoicing. Testing the graph showed the interface cut shared reads from 11 to 2, but the merge created a cycle with Tax. It dropped the merge. The architect chose the interface.

More agents for software architects

Decisions · IT and development

Blockchain Use Case Feasibility Screen Agent

A fair, documented screen of a blockchain proposal with a clear recommendation

10 steps2 checks6 jobs
Decisions · Risk management

Static Analysis Findings Triage Agent

Turn a long findings list into confirmed fixes and justified suppressions

10 steps2 checks3 jobs
Coding · Quality control

Test Coverage Gap Closing Agent

Add tests to the riskiest untested code and prove that they catch real bugs

10 steps2 checks11 jobs
Decisions · IT and development

Model Selection Bake-Off Agent

Choose a model with evidence on quality, cost and speed for the real task

11 steps2 checks7 jobs