obra

Explores user intent, requirements, and design considerations before implementation.

Evaluates code review feedback for clarity and technical merit before implementation.

Request peer review of code changes to verify quality and adherence to requirements.

Execute written implementation plans in separate sessions with review checkpoints.

Determines the best approach to integrate completed work into your main codebase.

Creates isolated workspaces for feature development using native git worktree or fallback methods.

Distribute independent tasks across multiple agents simultaneously without waiting for sequential completion.

Execute implementation plans by delegating independent tasks to specialized subagents within your session.

Methodically diagnoses bugs and test failures by analyzing root causes before proposing fixes.

Establishes skill invocation protocols and guides users on finding and accessing available capabilities before responding.

Run verification commands and confirm output before claiming work is complete or fixed.

Converts specifications into structured multi-step task plans before implementation begins.

Create, edit, and verify AI agent skills before deployment.