Endpoint readiness · Repo intelligence · Governed AI operations
I build a local AI operating system for infrastructure, endpoint readiness, repo intelligence, and safe automation.
- Run the client readiness tools — test the public endpoint-readiness surface.
- See how the system works — follow the signal, score, gate, and memory loop.
- Explore the terminal entrypoint — start with the front door to the MQ stack.
The MQ stack connects local repositories, endpoint operations, and AI-assisted engineering through one practical loop:
endpoint / repo / workflow → signal → score → gate → memory → better next action
The focus is operational: make state visible, decisions explainable, and automation safe enough to use under real pressure.
| Repository | Role |
|---|---|
macos-scripts |
Terminal entrypoint and local workflow toolkit |
mq-agent |
Orchestrates sweeps, reviews, release gates, and alerts |
mq-mcp |
Policy-bound MCP runtime for controlled tool execution |
repo-signal |
Scores repo readiness and exports structured AI context |
mq-image-analyze |
Extracts operational signal from screenshots and UI states |
mq-ums |
Provides a gated operator surface for IGEL UMS workflows |
One entrypoint, layered workflows, reviewable actions.
The model favors signal before action, local execution, explicit policy gates, and reusable technical memory. Read the operating model.
macos-scripts— the terminal front door for MQ workflows, diagnostics, and stack control.mq-agent— coordinates repo intelligence and operational workflows without hiding what happens.mq-mcp— makes AI tool use predictable through contracts, policy gates, and explicit boundaries.repo-signal— turns repository state into readiness scores, release checks, and agent context.