Focus.AI Labs / Layer Report
Subject: Observability and memory layer
Ecosystem layer landing page
Layer landing page

Once the agent matters, visibility matters too.

Saved chats, status bars, dashboards, and debug monitors show up quickly once users start trusting Cursor with real work and need logs, replay, and usage visibility.

Layer role Visibility, replay, and behavior inspection
Representative projects CursorLens, specstory, Cursor Stats
Navigation move Use this page to find the tools that explain what Cursor actually did, not just what it was asked to do.
Layer thesis

How to read this branch of the ecosystem

This is the shortest explanation of what this layer is for and why it matters.

Observability is the practical sign that the ecosystem is maturing from novelty to operations.

Key artifacts

Best entry points in this layer

These projects explain this branch of the ecosystem fastest.

Primary project

CursorLens

An open-source dashboard for Cursor.sh IDE. Log AI code generations, track usage, and control AI models (including local ones). Run locally or use upcoming hosted version.

Key artifact

specstory

SpecStory automatically saves every Cursor chat and composer session to your local project's .specstory directory.

Key artifact

Chrome Debug Monitor

A powerful integration between Chrome's DevTools Protocol and Cursor Composer for real-time debugging and monitoring.

Work patterns

What people actually do here

These workflow slices connect the layer to real usage patterns.

Workflow pattern

Instrument the agent once the work gets real

Logs, saved sessions, stats bars, dashboards, and browser-debug surfaces show up when teams start trusting Cursor for real development work and need to understand behavior, cost, and history.

Guardrails

What to remember while exploring

Layer-specific best-practice reminders that keep this branch legible.

Best practice

Add observability before you add more autonomy

The presence of chat archives, usage stats, dashboards, and debugging tools suggests a practical pattern: once Cursor starts doing real work, users want logging, replay, and visibility before they scale complexity further.