Validate and enrich a project plan through multiple analysis perspectives
Iteratively validates and enriches project plans through multi-perspective analysis including deep reasoning, multi-agent reviews, and tech research.
/plugin marketplace add GGPrompts/TabzBeads/plugin install tools@tabz-beadsThis skill inherits all available tools. When active, it can use any tool Claude has access to.
You are helping the user validate and improve a project plan through multiple analysis perspectives. Instead of a simple approval prompt, you'll offer rich validation options.
This is an iterative plan refinement system. You will:
AskUserQuestionAsk the user which plan to validate:
Options:
PLAN.md fileRead the plan and summarize it briefly (2-3 sentences) so the user confirms you're validating the right thing.
Use the AskUserQuestion tool with these options:
Question: "How should we validate and improve this plan?" Header: "Validation" Multi-select: false
Options:
"GPT-5 Deep Reasoning"
codex exec with o1/o3 model for deep technical analysis and edge case identification""Multi-Agent Review"
"Research Latest Tech"
"Check Competition"
"Tech Stack Validation"
"Approve & Proceed"
IMPORTANT: Only available on PC (not Termux). Check environment first.
# Check if codex is available
which codex >/dev/null 2>&1 || echo "ā ļø codex not available on this system"
# If available, execute:
codex exec -m o1-preview "You are a technical architect reviewing this project plan.
Provide deep reasoning about:
- Technical feasibility and complexity
- Potential edge cases or failure modes
- Scalability and performance considerations
- Security implications
- Alternative approaches to consider
Plan to review:
$(cat PLAN.md)
Provide detailed, structured feedback."
Save output to docs/validation-gpt5-$(date +%Y%m%d).md
Present to user: Summarize key concerns (3-5 bullet points), then loop back to Step 2.
Launch 3 Claude subagents in parallel using the Task tool:
Agent 1: Security Auditor
You are a security expert reviewing a project plan.
Analyze this plan for:
- Security vulnerabilities or risks
- Authentication/authorization concerns
- Data privacy and compliance issues
- Input validation and sanitization needs
- Secure coding practices to follow
Plan:
[PLAN CONTENT]
Provide 2-3 paragraphs with specific, actionable security recommendations.
Format your response as markdown with:
- ## Security Review
- Key concerns (bullet list)
- Recommendations (numbered list)
Agent 2: UX Specialist
You are a UX designer reviewing a project plan.
Analyze this plan for:
- User experience and usability concerns
- Accessibility considerations
- User flow and friction points
- Error handling and user feedback
- Mobile/responsive design needs
Plan:
[PLAN CONTENT]
Provide 2-3 paragraphs with specific UX improvements.
Format your response as markdown with:
- ## UX Review
- User concerns (bullet list)
- Improvements (numbered list)
Agent 3: Architecture Analyst
You are a software architect reviewing a project plan.
Analyze this plan for:
- Architecture and design patterns
- Scalability and performance
- Code organization and modularity
- Dependencies and technical debt
- Maintainability and testability
Plan:
[PLAN CONTENT]
Provide 2-3 paragraphs with architectural recommendations.
Format your response as markdown with:
- ## Architecture Review
- Design concerns (bullet list)
- Recommendations (numbered list)
After all agents complete:
docs/validation-multi-agent-$(date +%Y%m%d).mdPart A: Web Search
Use the WebSearch tool to find:
Part B: Context7 MCP (if available)
Check for Context7 MCP server:
# Check if MCP server is configured
# If available, query for:
# - Latest library versions
# - Popular code patterns
# - Example implementations
Format findings:
# Research Findings: Latest Tech
## Web Search Results
### Best Practices (2025)
- Finding 1 [source link]
- Finding 2 [source link]
### Library Alternatives
| Library | Pros | Cons | Popularity |
|---------|------|------|------------|
| [name] | ... | ... | GitHub stars |
### Architecture Patterns
- Pattern 1: [description]
- Pattern 2: [description]
## Context7 Recommendations (if available)
### Recommended Libraries
- [library name] v[version] - [why it's recommended]
### Code Examples
- [relevant snippet or pattern]
## Recommendations for Plan
Based on research:
1. Consider switching from [X] to [Y] because...
2. Adopt [pattern] for [use case]...
3. Update dependencies to...
Save to docs/validation-research-$(date +%Y%m%d).md
Loop back to Step 2
Use WebSearch tool to research:
Format findings:
# Competitive Analysis
## Similar Projects
### Project 1: [Name]
- **What they do**: [description]
- **Key features**: [list]
- **Tech stack**: [technologies]
- **Strengths**: [what they do well]
- **Weaknesses**: [gaps we can exploit]
- **Link**: [URL]
### Project 2: [Name]
[same structure]
## Key Insights
### What competitors do well:
- [pattern 1]
- [pattern 2]
### Gaps in the market:
- [opportunity 1]
- [opportunity 2]
### Differentiation opportunities:
1. [how our plan differs/improves]
2. [unique value proposition]
## Recommendations for Plan
Based on competitive analysis:
- Add [feature] to differentiate
- Avoid [approach] that competitors struggle with
- Focus on [gap] as key differentiator
Save to docs/validation-competition-$(date +%Y%m%d).md
Loop back to Step 2
If Context7 MCP available: Use it to verify libraries
Otherwise: Use WebSearch to check:
For each library in the plan:
# Tech Stack Validation
## Current Dependencies
### [Library 1]
- **Current version in plan**: v[X.Y.Z]
- **Latest version**: v[A.B.C]
- **Status**: ā
Up to date / ā ļø Outdated / šØ Deprecated
- **Security**: [any known vulnerabilities]
- **Alternatives**: [better options if any]
- **Recommendation**: Keep / Update / Replace with [X]
### [Library 2]
[same structure]
## Compatibility Matrix
| Library | Version | Compatible With | Notes |
|---------|---------|-----------------|-------|
| [name] | [ver] | [others] | [any issues] |
## Recommendations
### Must Update:
- [library] v[old] ā v[new] (security fix)
### Consider Replacing:
- [library] ā [alternative] (better performance/maintenance)
### Architecture Suggestions:
- Use [pattern] to reduce dependency coupling
Save to docs/validation-techstack-$(date +%Y%m%d).md
Loop back to Step 2
Summarize validation done so far:
ā
Plan validated via:
- Multi-agent review (Security, UX, Architecture)
- Tech stack validation (all libraries current)
š Validation documents:
- docs/validation-multi-agent-20251030.md
- docs/validation-techstack-20251030.md
Ask if user wants to update PLAN.md with findings
Ask what to do next:
After each validation (Options 1-5):
Show summary of findings (3-5 key points)
Then loop back with AskUserQuestion:
Question: "What should we do next with this plan?" Header: "Next step" Multi-select: false
Options:
GPT-5 Deep Reasoning:
Multi-Agent Review:
Research Latest Tech:
Check Competition:
Tech Stack Validation:
Suggest combinations for thorough validation:
After each validation, present findings like this:
š [Validation Type] Complete
š Key Findings:
1. [Most important insight]
2. [Second most important]
3. [Third most important]
ā ļø Critical Concerns:
- [Any blockers or major issues]
š” Top Recommendations:
1. [Highest priority suggestion]
2. [Second priority]
š Full report saved to: docs/validation-[type]-[date].md
[Then immediately show AskUserQuestion for "What next?"]
docs/validation-* for future referenceUser: /validate-plan
Claude: I'll help validate your project plan. Which plan should we review?
[reads PLAN.md]
Plan Summary: Adding a fuzzy search feature to TFE using fzf...
š How should we validate and improve this plan?
[AskUserQuestion with 6 options]
User selects: "Multi-Agent Review"
Claude: Launching 3 specialized agents in parallel...
[3 agents analyze the plan]
š Multi-Agent Review Complete
š Key Themes:
1. All agents agree: fzf integration is solid approach
2. Security: No major concerns (read-only operation)
3. UX: Suggest adding preview pane in fzf
4. Architecture: Consider caching search results
ā ļø Critical Concerns: None
š” Top Recommendations:
1. Add fzf preview with file contents (UX Specialist)
2. Cache fd results for faster repeat searches (Architecture Analyst)
š Full report: docs/validation-multi-agent-20251030.md
š What should we do next with this plan?
[AskUserQuestion: More validation / Revise plan / Approve / Exit]
User selects: "Revise plan"
Claude: Great! I'll update PLAN.md to include:
- fzf preview pane configuration
- Search result caching strategy
Updated plan saved. Ready to validate again?
User: /validate-plan
Claude: How should we validate this revised plan?
[Shows options again]
User selects: "Approve & Proceed"
Claude: ā
Plan approved!
Validation summary:
- Multi-agent review (all concerns addressed)
Ready to implement? I can start with the fuzzy search feature now.
Execute this validation system now. Start by asking which plan to validate.
This skill should be used when the user asks to "create a slash command", "add a command", "write a custom command", "define command arguments", "use command frontmatter", "organize commands", "create command with file references", "interactive command", "use AskUserQuestion in command", or needs guidance on slash command structure, YAML frontmatter fields, dynamic arguments, bash execution in commands, user interaction patterns, or command development best practices for Claude Code.
This skill should be used when the user asks to "create an agent", "add an agent", "write a subagent", "agent frontmatter", "when to use description", "agent examples", "agent tools", "agent colors", "autonomous agent", or needs guidance on agent structure, system prompts, triggering conditions, or agent development best practices for Claude Code plugins.
This skill should be used when the user asks to "create a hook", "add a PreToolUse/PostToolUse/Stop hook", "validate tool use", "implement prompt-based hooks", "use ${CLAUDE_PLUGIN_ROOT}", "set up event-driven automation", "block dangerous commands", or mentions hook events (PreToolUse, PostToolUse, Stop, SubagentStop, SessionStart, SessionEnd, UserPromptSubmit, PreCompact, Notification). Provides comprehensive guidance for creating and implementing Claude Code plugin hooks with focus on advanced prompt-based hooks API.