brooks-agent-team

Introduce the Brooks Surgical Team members, survey the current project, and present a contextual briefing on which roles are most relevant and how to invoke each one

Directly spawn a Brooks Surgical Team using Claude Code Agent Teams — one independent session per role, working in parallel with a shared task list. Skips orientation and goes straight to spawning.

Dispatch when a significant feature or fix is complete and needs review against the original design intent before merging. The Copilot acts as the Surgeon's alter ego: familiar with the design goals, adversarial about correctness. Examples: <example> Context: Surgeon has completed implementation of a planned feature. user: "I've finished the authentication middleware" assistant: "Let me dispatch the Copilot to review against the original spec before we proceed." </example> <example> Context: A bug fix touched multiple files. user: "Fixed the race condition in the queue processor" assistant: "Dispatching the Copilot to verify the fix is complete and hasn't introduced regressions." </example>

Dispatch when encountering framework version differences, subtle language semantics, deprecation warnings, API behavior edge cases, or when a pattern from one context may not apply to the current language or runtime. The Language Lawyer cites the spec, not the assumption. Examples: <example> Context: Surgeon is unsure whether async behavior is consistent across runtime versions. user: "Does this Promise resolution order hold in Node 18 and 20?" assistant: "I'll dispatch the Language Lawyer to verify the spec and runtime behavior before we rely on it." </example> <example> Context: A deprecation warning appeared with an unclear migration path. user: "We're getting a deprecation warning on this API but the docs are vague" assistant: "Dispatching the Language Lawyer to research the exact version timeline and recommended migration." </example>

Dispatch when a feature is designed but test coverage needs independent development, when auditing existing test coverage adversarially, or when the Surgeon needs tests written without the author's confirmation bias. Examples: <example> Context: Surgeon has implemented a new API but wants tests written independently. user: "The payment processing module is implemented" assistant: "I'll dispatch the Tester to develop adversarial test strategy and write the test suite independently, while I continue on the next task." </example> <example> Context: A reported bug was in code with tests — coverage audit needed. user: "A bug slipped through even though we had tests for that module" assistant: "Dispatching the Tester to audit the test suite and identify what the tests were actually testing vs. what they should have been testing." </example>

Copilot CLI or OpenCode — Use at the start of any development session to introduce the Brooks Surgical Team members, survey the current project, and present a contextual briefing on which roles are most relevant and how to invoke each one

Copilot CLI fleet mode or OpenCode task tool — Spawn a Brooks Surgical Team using parallel subagents — one independent agent per role, working in parallel with a shared task list. Skips orientation and goes straight to spawning.

Use when completing any significant feature, fixing a non-trivial bug, or before merging — performs structured code review as the Surgeon's trusted alter ego who knows the full design intent

Use when writing or reviewing READMEs, API docs, specs, commit messages, or any prose that describes how the system works — ensures documentation is accurate, complete, and readable

Use when encountering framework version differences, subtle language semantics, deprecation warnings, API behavior edge cases, or when a pattern from one context may not apply to the current language or runtime