The problem

Local Windows AI projects often fail due to port conflicts, broken Python virtual environments, or mismatched Ollama configurations. Developers waste hours debugging "Unexpected token <" errors and cryptic BAT script failures that close windows before logs can be read.

What it does

• Identifies blocked ports and the specific PIDs holding them using PowerShell diagnostics.
• Validates Python environments and ensures the correct interpreter and dependencies are active.
• Tests Ollama API health and verifies model availability, tags, and connection timeouts.
• Inspects raw API responses to fix common "HTML-instead-of-JSON" frontend failures.
• Audits BAT and PowerShell launchers for path errors, drive mismatches, and improper command chaining.

Frameworks & tools

Windows 10/11, Python (Flask, FastAPI), Node.js, Ollama, PowerShell, and BAT scripts.

Why this beats prompting it yourself

A standard LLM often suggests destructive fixes like reinstalling everything. This skill follows a strict "diagnose before modify" protocol, creating reversible backups and verifying network bindings before touching a single line of code.

Use cases

• Recovering a local AI dashboard that fails to connect to the backend after a Windows update.
• Fixing "model not found" or 404 errors in Ollama-based applications.
• Debugging Python startup failures where the terminal window closes too fast to see errors.
• Resolving CORS or port binding issues in Flask/FastAPI and Node.js stacks.

Known limitations

Designed specifically for Windows environments. It does not perform destructive actions like firewall rule changes or package installs without explicit user confirmation.