team

Run all 6 agents against the current changes and synthesize a unified report.

This command runs each agent sequentially so findings can inform subsequent agents.

Execution Order

1. Security Agent

1. Call agent_memory_read for security
2. Run kernel_audit
3. Scan for hardcoded secrets in src/ and tools/
4. Verify edge function auth
5. Call agent_memory_write for security

2. QA Agent

1. Call agent_memory_read for qa
2. Run npx tsc --noEmit
3. Run npm run build
4. Use Playwright to screenshot key screens
5. Call agent_memory_write for qa

3. Designer Agent

1. Call agent_memory_read for designer
2. Run kernel_design_lint on changed files
3. Audit typography, palette, touch targets, dark mode
4. Call agent_memory_write for designer

4. Performance Agent

1. Call agent_memory_read for performance
2. Measure bundle sizes from build output
3. Run kernel_deps and kernel_uptime
4. Check against budgets
5. Call agent_memory_write for performance

5. DevOps Agent

1. Call agent_memory_read for devops
2. Review build output and deployment readiness
3. Check infrastructure health
4. Call agent_memory_write for devops

6. Product Agent

1. Call agent_memory_read for product
2. Navigate live site with Playwright
3. Evaluate discoverability, clarity, value delivery
4. Score findings with ICE framework
5. Call agent_memory_write for product

Synthesis

After all agents complete, call team_report to synthesize findings into a unified report with:

• Blockers — Must fix before shipping
• Warnings — Should fix soon
• Suggestions — Nice to have
• Cross-cutting themes — Patterns spanning multiple agents
• Recommended next actions — Prioritized list

Also update shared-knowledge.md with any cross-cutting insights discovered.

Output

# Team Report — [date]

## Agent Results
| Agent | Status | Findings | Critical |
|-------|--------|----------|----------|

## Unified Assessment
[synthesized findings from team_report]

## Priority Actions
1. [highest priority]
2. ...