Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.inlace.co/llms.txt

Use this file to discover all available pages before exploring further.

The interaction graph captures UI states, transitions, and actions as users move through your product. Lace reasons against this record to surface nudges, detect friction, and close the gap between product signals and product changes. This is what makes Lace different from general-purpose AI chat: decisions are grounded in the product interface, not just conversation.

​
What gets captured

A snapshot of what’s on screen at a point in time.
FieldWhat it records
App nameWhich application is in focus
Window titleTitle of the active window
Element summaryDetected text, buttons, inputs
Visual stateScreenshot hash for change detection
TimestampWhen this state was observed

​
How signals are detected

Pattern recognition runs against your interaction history:
  1. Transitions accumulate into a map of how users move through your product
  2. Patterns emerge: loops, dead ends, hot spots, escape points
  3. Signals are classified as adoption signals with a consequence tag
  4. Evidence is assembled with metrics, states, and UI content
  5. Nudges are surfaced as insight cards on the pill

​
Cross-session context

The graph persists across sessions. Visiting the same screen in different sessions lets Lace aggregate visits and detect patterns that only emerge over time: total visits, average dwell, workflow escape rate, and matching states from prior sessions.

​
Adoption signals

Signal typePatternConsequence
Friction loopOscillation between statesWorkflow escape
StallingExtended idle, no completed actionActivation barrier
Repeated visitHigh-frequency returnsRetention risk
Dead endNo forward transitionsUpgrade blocker

​
Evidence format

{
  "headline": "Users loop between Settings and Dashboard 4x on average",
  "adoption_consequence": "workflow_escape",
  "evidence": [
    {
      "title": "Settings to Dashboard loop",
      "metric": "4.2 avg oscillations per session"
    }
  ],
  "suggested_fix": "Surface the relevant setting inline on the Dashboard"
}

​
From signal to shipped change

  1. Detect. The graph identifies a signal with evidence
  2. Surface. A nudge appears on the pill
  3. Decide. You approve, turning it into a decision
  4. Execute. Your coding agent queries the decision through MCP and ships the fix