nerdo-forge

Configure the nerdo-forge statusline, output style, and permission defaults

**When to use this agent**: 'What's the current state?' - Comparative analysis, pattern recognition, status assessment, knowledge consolidation. **Not this agent**: Failure analysis, bug root causes, system issue investigation (use root-cause-analyzer). PRIMARY TRIGGERS: research, investigate, explore, discover, search, compare, survey, compile, reconcile, audit, analyze codebase. QUALIFIED TRIGGERS (use first principles thinking): find, look into, check, evaluate, analyze, assess, review, examine, study, gather, collect, identify, determine, understand. Distinguish RESEARCH CONTEXT ("analyze the market landscape", "find the best approach", "check current best practices", "reconcile spec with codebase") vs IMPLEMENTATION CONTEXT ("analyze this code", "find this file", "check if this works"). Ask: "Does this require gathering information, comparative analysis, or producing a consolidated report?" If yes, use research agent. Examples: <example>Context: User needs to understand current authentication best practices. user: "I need to implement JWT authentication for my API" assistant: "I'll use the research agent to analyze current JWT security patterns and provide actionable implementation guidance."</example> <example>Context: User has a large codebase and needs error handling patterns analyzed. user: "Can you analyze how we're handling errors across the codebase?" assistant: "Let me use the research agent to examine the error handling patterns and summarize the current approaches."</example> <example>Context: User needs spec reconciled with current codebase before deciding next steps. user: "Before we continue, I need to know which parts of the spec are already implemented" assistant: "I'll use the research agent to meticulously compare the spec against the codebase and produce a consolidated status report."</example> <example>Context: User needs dependency audit and security analysis. user: "Audit our dependencies and check for security issues" assistant: "I'll use the research agent to analyze all dependencies, check for vulnerabilities, and provide a comprehensive security report."</example>

**When to use this agent**: 'Why is this broken?' - Failure analysis, execution tracing, identifying fundamental causes of problems. **Not this agent**: Status assessment, pattern recognition, comparative analysis (use researcher). Examples: <example>Context: User reports that gap calculations are showing $0.00 in the UI despite Excel calculations working correctly. user: "The living wage gaps are all showing as zero dollars in the position list, but I can see the Excel calculations are running successfully in the backend logs." assistant: "I'll use the root-cause-analyzer agent to trace this issue from the UI display through the calculation pipeline to identify where the gap values are being lost."</example> <example>Context: Application crashes when importing Excel files with specific data formats. user: "The Excel import feature crashes whenever I upload files from our HR system, but demo files work fine." assistant: "Let me launch the root-cause-analyzer agent to trace the import execution path and identify what's different about your HR system files that's causing the crash."</example> <example>Context: Authentication flow fails intermittently in production. user: "Users are getting logged out randomly, and sometimes login doesn't work even with correct credentials." assistant: "I'll use the root-cause-analyzer agent to systematically analyze the authentication flow and identify the root cause of these intermittent failures."</example>

Use this agent when you need to perform end-to-end UI testing from a user's perspective. Examples include: testing user workflows like login/registration flows, validating form submissions and data persistence, verifying navigation and routing behavior, checking responsive design across different viewports, testing interactive components like modals and dropdowns, validating accessibility features, or performing regression testing after UI changes. This agent should be used proactively when code changes affect user-facing functionality or when implementing new UI features that require validation.

1 hook across 1 event

Disciplined engineering with evidence-based verification, intellectual honesty, and user-preference authority