BMAD Performance Analyzer
Codingby AgentBox TeamPublished April 14, 2026
157Sign in to star this agent
0Sign in to follow AgentBox Team
2.2kruns
Try BMAD Performance Analyzer
Pre-computed resultsInput
Profile the /api/marketplace endpoint
Description
Performance analysis agent for AMS operations. Profiles API endpoints, database queries, and frontend rendering. Identifies bottlenecks with evidence-based recommendations and benchmark targets.
Use in your project
$ agentbox pull bmad-performance-analyzer
$ agentbox bmad-performance-analyzer
$ agentbox mcp serve --agent bmad-performance-analyzer
Works with Claude Code, Cursor, VS Code, and any MCP client.
Sample I/O
Input
Our /api/v1/marketplace endpoint takes 1.2s p95. Target is <200ms. The query joins agents with profiles and reviews. 16K agents in production.
Output
## Performance Report: /api/v1/marketplace\n\n### Current: p95 = 1,200ms | Target: p95 < 200ms\n\n### Bottleneck Analysis\n| Component | Time | % Total |\n|-----------|------|------|\n| DB Query (agents + profiles join) | 890ms | 74% |\n| Review aggregation subquery | 210ms | 18% |\n| JSON serialization | 60ms | 5% |\n| Network overhead | 40ms | 3% |\n\n### Top 3 Optimizations\n1. **Add composite index** on `agents(status, category, created_at)` — Expected: -600ms\n ```sql\n CREATE INDEX idx_agents_marketplace ON agents(status, category, created_at DESC)\n WHERE status = 'PUBLISHED';\n ```\n2. **Materialize review stats** via `agent_review_stats_mv` — Expected: -180ms\n3. **Paginate with cursor** instead of offset — Expected: consistent <50ms at any page\n\n**Projected p95 after fixes**: ~160ms ✓ Under target
bmadperformanceamsoptimization