/team-review — Parallel Subagent Code Review

Spawn 4 review subagents simultaneously, each with its own context window.

Execution

1. •Identify scope: git diff --name-only HEAD~1 (or specified scope)
2. •Read each agent definition from .claude/agents/
3. •Spawn 4 Task tool calls in the same response for parallel execution:

Task 1 — Security:

Task(subagent_type="security-code-auditor",
     prompt="<content of .claude/agents/security-reviewer.md>\n\nReview these files:\n{file_list}")

Task 2 — Architecture:

Task(subagent_type="senior-code-reviewer",
     prompt="<content of .claude/agents/architecture-reviewer.md>\n\nReview these files:\n{file_list}")

Task 3 — Code Quality:

Task(subagent_type="senior-code-reviewer",
     prompt="<content of .claude/agents/code-reviewer.md>\n\nReview these files:\n{file_list}")

Task 4 — Test Quality:

Task(subagent_type="senior-code-reviewer",
     prompt="<content of .claude/agents/test-reviewer.md>\n\nReview these files:\n{file_list}")

1. •Collect all 4 results
2. •Synthesize into unified report sorted by severity (Critical > Major > Minor)

Arguments

• •/team-review — review uncommitted changes
• •/team-review crates/ha-api — review specific crate

Why Parallel?

Each reviewer has its own context window — no context dilution between dimensions. All 4 run simultaneously via Task tool.