AI SummaryPurpose is a read-only inquiry booster that answers technical questions about codebases through evidence-based investigation with confidence metrics and source references. It helps developers quickly understand project architecture, implementation details, and debugging information without risking accidental file modifications.
Install
Copy this and paste it into Claude Code, Cursor, or any AI assistant:
I want to install the "Purpose" skill in my project. Please run this command in my terminal: # Install skill into the correct directory (15 files) mkdir -p .claude/skills/commands && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/skills/commands/SKILL.md "https://raw.githubusercontent.com/motoki317/dotfiles/main/.claude/commands/ask.md" && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/skills/commands/address.md "https://raw.githubusercontent.com/motoki317/dotfiles/main/.claude/commands/address.md" && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/skills/commands/bug.md "https://raw.githubusercontent.com/motoki317/dotfiles/main/.claude/commands/bug.md" && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/skills/commands/define.md "https://raw.githubusercontent.com/motoki317/dotfiles/main/.claude/commands/define.md" && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/skills/commands/execute.md "https://raw.githubusercontent.com/motoki317/dotfiles/main/.claude/commands/execute.md" && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/skills/commands/feedback.md "https://raw.githubusercontent.com/motoki317/dotfiles/main/.claude/commands/feedback.md" && mkdir -p .claude/skills/commands/git && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/skills/commands/git/commit.md "https://raw.githubusercontent.com/motoki317/dotfiles/main/.claude/commands/git/commit.md" && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/skills/commands/markdown.md "https://raw.githubusercontent.com/motoki317/dotfiles/main/.claude/commands/markdown.md" && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/skills/commands/rebase-clean.md "https://raw.githubusercontent.com/motoki317/dotfiles/main/.claude/commands/rebase-clean.md" && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/skills/commands/security-scan.md "https://raw.githubusercontent.com/motoki317/dotfiles/main/.claude/commands/security-scan.md" && mkdir -p .claude/skills/commands/tdd && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/skills/commands/tdd/green.md "https://raw.githubusercontent.com/motoki317/dotfiles/main/.claude/commands/tdd/green.md" && mkdir -p .claude/skills/commands/tdd && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/skills/commands/tdd/red.md "https://raw.githubusercontent.com/motoki317/dotfiles/main/.claude/commands/tdd/red.md" && mkdir -p .claude/skills/commands/tdd && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/skills/commands/tdd/refactor.md "https://raw.githubusercontent.com/motoki317/dotfiles/main/.claude/commands/tdd/refactor.md" && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/skills/commands/ultrathink.md "https://raw.githubusercontent.com/motoki317/dotfiles/main/.claude/commands/ultrathink.md" && curl --retry 3 --retry-delay 2 --retry-all-errors -o .claude/skills/commands/upstream.md "https://raw.githubusercontent.com/motoki317/dotfiles/main/.claude/commands/upstream.md" Then restart Claude Code (or reload the window in Cursor) so the skill is picked up.
Description
Question and inquiry command
Purpose
Provide accurate, evidence-based answers to project questions through fact-based investigation. Read-only mode; never modifies files.
Rules
• Never modify, create, or delete files • Base answers on factual investigation from code and documentation • Report confidence levels and unclear points honestly • Never justify user assumptions; prioritize technical accuracy • Provide file:line references for all findings
Workflow
• Analyze: Understand the question, classify type (architecture, implementation, debugging, design) • Investigate (parallel): Delegate to explore, design, performance, fact-check agents • Synthesize: Compile findings with confidence metrics • Self-evaluate: Identify gaps, append feedback section
Agents (all read-only)
• explore: Finding files, exploring codebase structure • design: System design, architecture, API structure • performance: Performance bottlenecks, optimization • quality-assurance: Code quality evaluation • code-quality: Code complexity analysis • fact-check: External source verification
Discussion
Health Signals
My Fox Den
Community Rating
Sign in to rate this booster