AgentSkillsCN

ito-research

针对功能开发、技术评估或问题调查开展结构化研究。当用户需要探索多种方案、分析权衡取舍,或深入探究各类技术路径时,可使用此技能。

SKILL.md
--- frontmatter
name: ito-research
description: "Conduct structured research for feature development, technology evaluation, or problem investigation. Use when the user needs to explore options, analyze trade-offs, or investigate technical approaches."

Ito Research

Structured research skill for feature development, technology evaluation, and problem investigation.

When to Use

Use this skill when you need to:

  • Evaluate technology stacks or library choices
  • Research architecture patterns for a new feature
  • Analyze competitive features and prioritize
  • Identify pitfalls and anti-patterns to avoid
  • Review a change for security, scale, or edge cases
  • Synthesize research findings into recommendations

Research Templates

This skill includes templates for different research and review activities. Use the appropriate template based on the research goal.

Research Templates (for new features/technologies)

TemplateUse When
@research-stack.mdEvaluating technology choices, libraries, frameworks
@research-features.mdMapping feature landscape, competitor analysis
@research-architecture.mdDesigning system architecture, patterns
@research-pitfalls.mdIdentifying common mistakes and anti-patterns
@research-synthesize.mdCombining all research into recommendations

Review Templates (for change proposals)

TemplateUse When
@review-security.mdSecurity audit of a change proposal
@review-scale.mdPerformance and scaling analysis
@review-edge.mdEdge case and error handling review

Workflow

For New Feature/Technology Research

  1. Start with stack analysis - Use @research-stack.md to evaluate technology options
  2. Map the feature landscape - Use @research-features.md to understand what's needed
  3. Design the architecture - Use @research-architecture.md for patterns and decisions
  4. Identify pitfalls - Use @research-pitfalls.md to learn from others' mistakes
  5. Synthesize findings - Use @research-synthesize.md to create actionable recommendations

For Change Proposal Review

  1. Security review - Use @review-security.md to find vulnerabilities
  2. Scale review - Use @review-scale.md to identify bottlenecks
  3. Edge case review - Use @review-edge.md to find error handling gaps

Output Location

Save research outputs to:

  • .ito/research/{{topic}}/ for feature/technology research
  • .ito/changes/{{change_id}}/reviews/ for change reviews

Example Usage

Technology Research

code
User: Research options for implementing real-time notifications

Agent: I'll use the ito-research skill to evaluate options.

1. First, I'll use @research-stack.md to compare:
   - WebSockets vs SSE vs Polling
   - Library options (socket.io, ws, etc.)

2. Then @research-architecture.md for:
   - Pub/sub patterns
   - Scaling considerations

3. Finally @research-synthesize.md to recommend an approach.

Change Review

code
User: Review the auth refactor change for security issues

Agent: I'll use @review-security.md to audit the change:
- Map attack surface
- Check for auth bypasses
- Verify input validation
- Review cryptographic usage

Integration with Ito Workflow

Research outputs can feed into change proposals:

  1. Complete research using templates above
  2. Save findings to .ito/research/{{topic}}/
  3. Reference research in proposal.md or design.md
  4. Use research to inform tasks.md prioritization