UX Performance Sweep: Low-Risk High-Reward Snappiness Optimizations #371
Description
Summary
Systematic sweep through the codebase for low-risk, high-reward UX performance boosts. Targets React rendering efficiency, main process responsiveness, and database query performance.
Playbook Structure
8 auto run documents in Auto Run Docs/2026-02-14-UX-Perf-Sweep/:
| Doc | Focus | Tasks |
|---|---|---|
| PERF-01 | React.memo wraps for list items & high-churn components |
7 |
| PERF-02 | Memoize expensive render-path computations (sorts, regex, dates) | 7 |
| PERF-03 | Extract inline styles to memoized constants | 4 |
| PERF-04 | Main process async: history-manager.ts (biggest UI blocker) |
4 |
| PERF-05 | Main process async: stats-db, debug collectors, parsers | 5 |
| PERF-06 | DB compound indexes, IPC parallelization, agent session scanning | 4 |
| PERF-07 | 🔮 Seed doc — populated by PERF-06 with DB query analysis | 0 |
| PERF-08 | 🔮 Seed doc — populated by PERF-02/03 with discovered opportunities | 0 |
Meta-Strategy
PERF-07 and PERF-08 are seed documents that start nearly empty. Earlier docs carry "discovery tasks" that instruct the executing agent to write new tasks into these blank docs based on what they find:
- PERF-02 → discovers additional memoization opportunities → writes to PERF-08
- PERF-03 → discovers deeper re-render issues during style extraction → writes to PERF-08
- PERF-06 → runs
EXPLAIN QUERY PLANanalysis → writes verification tasks to PERF-07
Risk
All changes are low-risk:
React.memowraps can't break behavioruseMemofor sorts/regex is pure computation caching- Style extraction is cosmetic refactoring
- Async fs conversions follow established patterns
- DB indexes only speed up reads
Test Plan
- Run full playbook via Symphony end-to-end
- Verify
npm run lintpasses after all phases - Verify
npm run buildsucceeds - Manual snappiness check on session list, tab switching, history panel
PERF-01.md
PERF-02.md
PERF-03.md
PERF-04.md
PERF-05.md
PERF-06.md
PERF-07.md
PERF-08.md