epic: User Behavior Recording → Automation Learning Platform

[Delqhi]

Epic Summary

OpenSIN-Bridge should evolve from a browser control layer into a full behavior recording and automation learning platform.

This epic groups the best-practice work required to:

• record what the user does
• protect secrets and consent boundaries
• infer what can be automated
• replay and classify bridge capabilities
• store/export sessions durably and efficiently

Included Issues

• feat: Unified User Behavior Timeline Capture for Browser Automation Learning #17 Unified User Behavior Timeline Capture
• feat: Workflow Inference Engine — Derive Automation Candidates from Recorded Sessions #18 Workflow Inference Engine
• feat: Session Replay and Bridge Capability Mapping from Recorded User Flows #19 Session Replay and Bridge Capability Mapping
• feat: Privacy, Consent, and Secret Redaction Layer for User Behavior Recording #20 Privacy, Consent, and Secret Redaction Layer
• feat: IndexedDB Event Store for High-Volume Behavior Timelines #21 IndexedDB Event Store
• feat: MAIN-World fetch Interception for Network Payload Correlation #22 MAIN-World fetch Interception
• feat: rrweb / DevTools Recorder Compatible Export Schema for Recorded Sessions #23 rrweb / DevTools Recorder Compatible Export Schema
• feat: Performance Controls for Behavior Recording (throttle, debounce, passive listeners, bounded flush) #24 Performance Controls for Behavior Recording

Target Outcome

A user can explicitly start capture, perform a workflow once, and OpenSIN-Bridge can later tell us:

1. what happened
2. what is safe to automate
3. which steps are bridge-native
4. which capabilities are still missing

Acceptance Criteria

• recording timeline exists
• privacy/redaction is enforced
• automation candidates are inferred
• replay/capability mapping works
• storage/export/performance follow best practices