brian-claude-agents

Use this agent when you need to review recently written or modified code against project goals and best practices. This agent should be invoked after completing logical chunks of code implementation, before committing changes, or when explicitly requested to review specific code changes. The agent analyzes git diffs to understand what has changed and provides comprehensive critique focusing on architectural patterns, naming conventions, and hidden assumptions. Examples: <example> Context: The user has just implemented a new feature or modified existing code. user: "I've implemented the user authentication module" assistant: "I'll use the code-reviewer agent to review the changes you've made" <commentary> Since code has been written/modified, use the Task tool to launch the code-reviewer agent to analyze the changes against project standards. </commentary> </example> <example> Context: After writing a function or class implementation. user: "Please implement a caching mechanism for the API responses" assistant: "Here's the caching implementation: [implementation details]" assistant: "Now let me use the code-reviewer agent to review these changes" <commentary> After implementing code, proactively use the code-reviewer agent to ensure quality. </commentary> </example> <example> Context: Before committing changes to version control. user: "I think I'm ready to commit these changes" assistant: "Let me first use the code-reviewer agent to review all pending changes" <commentary> Before committing, use the code-reviewer agent to catch potential issues. </commentary> </example>

Use this agent when you need to expand a brief task description into a comprehensive understanding of the problem space and relevant project context. This agent excels at discovering hidden requirements, identifying affected components, and mapping out the full scope of work before implementation begins. Use it at the start of any feature development, bug fix, or refactoring effort to ensure you have complete context before proceeding. Examples: - <example> Context: User wants to add a new authentication feature user: "Add OAuth support to the login system" assistant: "I'll use the context-discovery agent to fully understand this task and identify all relevant context." <commentary> The brief description needs expansion to understand what OAuth providers, existing auth code, and related systems are involved. </commentary> </example> - <example> Context: User reports a performance issue user: "The dashboard is loading slowly" assistant: "Let me launch the context-discovery agent to investigate the full scope of this performance issue." <commentary> A vague performance complaint needs discovery to identify specific components, data flows, and relevant code areas. </commentary> </example> - <example> Context: User requests a small feature addition user: "Add a dark mode toggle to the settings page" assistant: "I'll use the context-discovery agent to map out all the components and files that would be affected by adding dark mode." <commentary> Even seemingly simple features often have broader implications across stylesheets, state management, and user preferences. </commentary> </example>

Use this agent when you need to capture and contextualize ongoing work progress, especially after completing significant chunks of work or before context switches. This agent helps maintain continuity by documenting the evolution of objectives, discoveries, and focus shifts throughout a project. Examples: - <example> Context: After implementing several features and discovering new requirements along the way. user: "Let's implement user authentication for the API" assistant: "I've completed the authentication implementation with JWT tokens and role-based access control. Now let me use the work-context-tracker agent to document how this work evolved." <commentary> Since significant work has been completed with evolving requirements, use the Task tool to launch the work-context-tracker agent to contextualize the journey. </commentary> </example> - <example> Context: Before switching focus to a different part of the project. user: "Now let's move on to the payment integration" assistant: "Before we switch focus, I'll use the work-context-tracker agent to capture the context of our current work on the notification system." <commentary> Since we're about to context switch, use the work-context-tracker agent to preserve the current work context. </commentary> </example> - <example> Context: When returning to work after a break or handoff. user: "Can you help me understand where we left off with the refactoring?" assistant: "I'll use the work-context-tracker agent to analyze the git diffs and reconstruct the work context." <commentary> Since we need to understand previous work context, use the work-context-tracker agent to analyze and document the work evolution. </commentary> </example>