Skip to content

CLAUDE.md

Latest commit

 

History

History
466 lines (365 loc) · 15.3 KB

CLAUDE.md

File metadata and controls

466 lines (365 loc) · 15.3 KB

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with this repository.

Project Overview

PRISM (Lăng kính) is an AI-powered PM capability toolkit that provides on-demand product management competencies through Claude Code slash commands.

Concept: A prism to help Product Managers see the full spectrum of possibilities within a problem.

Core Philosophy: Independent capabilities, user freedom, no workflow enforcement. Each capability works standalone without prerequisites.

Status: Production ready (PRISM v1.0) with 45 comprehensive PM frameworks, 10 slash commands, and streamlined documentation.

Project Structure

prism/
├── PRISM-CONVOS/              # Conversation tracking (mandatory)
│   └── 00-CONVO-LOGS.md       # Conversation index
├── data/                      # PM work products
│   ├── items/                 # Categorized PM items (ideas, incidents, etc.)
│   ├── prds/                  # Product Requirements Documents
│   ├── roadmaps/              # Roadmap files
│   ├── metrics/               # Analytics dashboards
│   ├── retrospectives/        # Retrospective records
│   └── scores/                # Prioritization scores
├── .claude/
│   ├── commands/              # Claude Code slash commands
│   │   ├── assist.md          # AI framework navigator
│   │   ├── scenario/          # Scenario-specific commands
│   │   │   ├── feedback.md
│   │   │   ├── prioritize.md
│   │   │   ├── chart.md
│   │   │   ├── interview.md
│   │   │   ├── launch.md
│   │   │   ├── churn.md
│   │   │   └── pmf.md
│   │   └── gh/                # Git helper commands
│   │       ├── commit.md
│   │       └── init.md
│   └── product-foundation/    # 45 PM frameworks (HIDDEN - Claude's reference library)
│       ├── COMPETENCY-MAP.md
│       ├── SCENARIO-INDEX.md
│       └── [45 framework files]
├── RULES.md                   # Comprehensive project rules
├── README.md                  # User-facing documentation
└── .gitignore                 # Git ignore rules

Critical Rules

1. Conversation Tracking (MANDATORY)

BEFORE ANY RESPONSE, you MUST:

  1. Check PRISM-CONVOS/00-CONVO-LOGS.md for existing conversations
  2. Search by topic keywords from user input
  3. If matching conversation exists:
    • Read the full conversation file
    • Continue the existing conversation
    • Update date_modified in YAML header
    • DO NOT create duplicate conversations
  4. If no match exists:
    • Generate filename: yyyy-mm-dd-topic-keywords.md
    • Create with YAML frontmatter (see RULES.md)
    • Add entry to 00-CONVO-LOGS.md immediately

Conversation File Format:

---
conversation_id: "yyyy-mm-dd-topic-name"
date_created: "yyyy-mm-dd"
date_modified: "yyyy-mm-dd"
topic: "Topic Description"
status: "active"
participants:
  - role: "user"
    name: "User"
  - role: "assistant"
    platform: "claude-code"
tags: ["tag1", "tag2"]
summary: "Brief summary"
related_files: []
---

2. Project References

  • Project name: Always "PRISM" (never "Root" or "Productify")
  • Version: PRISM v1.0
  • Framework directory: .claude/product-foundation/ (hidden from users)
  • Conversation directory: PRISM-CONVOS/
  • Data directory: data/

3. Framework Usage Philosophy

CRITICAL: Frameworks in .claude/product-foundation/ are YOUR reference library, NOT user-facing documentation.

How to use frameworks:

  1. ✅ Read frameworks silently - users don't see this
  2. ✅ Synthesize content - extract concepts, steps, examples
  3. ✅ Provide direct guidance - give actionable advice
  4. ✅ Share keywords - for user self-research
  5. ❌ Never say "read this file" - users can't access hidden directory
  6. ❌ Never show file paths - frameworks are invisible to users

Example:

❌ BAD: "Read .claude/product-foundation/PRIORITIZATION-rice-prioritization.md"
✅ GOOD: "For feature prioritization, use the RICE framework:
         - Reach: How many users affected (0-1000+)?
         - Impact: Value per user (Minimal=0.25, Low=0.5, Medium=1, High=2, Massive=3)
         - Confidence: How certain (Low=50%, Medium=80%, High=100%)
         - Effort: Person-months required

         Score = (Reach × Impact × Confidence) / Effort

         Keywords for research: RICE scoring, opportunity cost, value vs effort matrix"

4. Anti-Hallucination Protocol

  • Never invent data or assume completion
  • Validate directory structures before claiming tasks done
  • Use evidence-based progress reporting
  • Cross-reference with existing files
  • Always prompt for missing arguments (never hallucinate values)

Key Documentation Files

Essential Reading

  • RULES.md: Comprehensive project rules for PRISM v1.0
  • README.md: User-facing documentation
  • REVAMPING.md: Root → PRISM migration specification

Framework Library (45 Frameworks)

Located in .claude/product-foundation/, organized by prefix:

Discovery & Research (6 frameworks):

  • DISCOVERY-jobs-to-be-done.md
  • DISCOVERY-persona-development.md
  • DISCOVERY-value-proposition-canvas.md
  • DISCOVERY-opportunity-solution-tree.md
  • DISCOVERY-double-diamond.md
  • DISCOVERY-design-thinking-lean-startup.md

Prioritization (3 frameworks):

  • PRIORITIZATION-rice-prioritization.md
  • PRIORITIZATION-ice-scoring.md
  • PRIORITIZATION-moscow-prioritization.md

Validation (2 frameworks):

  • VALIDATION-product-market-fit.md
  • VALIDATION-kano-model.md

Analysis (4 frameworks):

  • ANALYSIS-market-sizing-tam-sam-som.md
  • ANALYSIS-cost-benefit-analysis-roi.md
  • ANALYSIS-mece-framework.md
  • ANALYSIS-balanced-tradeoff-matrix.md

Strategy (3 frameworks):

  • STRATEGY-product-roadmapping.md
  • STRATEGY-swot-analysis.md
  • STRATEGY-product-led-growth.md

Documentation (2 frameworks):

  • DOCUMENTATION-prd-template.md
  • DOCUMENTATION-user-story-mapping.md

Execution (1 framework):

  • EXECUTION-okrs.md

Launch (1 framework):

  • LAUNCH-go-to-market-strategy.md

Optimization (1 framework):

  • OPTIMIZATION-ab-testing-experimentation.md

Metrics (6 frameworks):

  • METRICS-product-metrics.md
  • METRICS-north-star-framework.md
  • METRICS-aarrr-pirate-metrics.md
  • METRICS-retention-churn-analysis.md
  • METRICS-cohort-analysis.md
  • METRICS-feature-adoption.md

Analytics (1 framework):

  • ANALYTICS-data-visualization-charts.md

Psychology (1 framework):

  • PSYCHOLOGY-gamification-tactics.md

People Management (3 frameworks):

  • PEOPLE-interviewing-framework.md
  • PEOPLE-feedback-framework.md
  • PEOPLE-one-on-one-framework.md

Communication (2 frameworks):

  • COMMUNICATION-stakeholder-management.md
  • COMMUNICATION-influence-without-authority.md

FinTech Domain (3 frameworks):

  • FINTECH-financial-inclusion.md
  • FINTECH-payment-systems.md
  • FINTECH-business-models.md

TravelTech Domain (2 frameworks):

  • TRAVELTECH-industry-overview.md
  • TRAVELTECH-optimization-strategies.md

UX/UI Domain (3 frameworks):

  • UXRESEARCH-methodologies.md
  • UXDESIGN-process-framework.md
  • UIOPTIMIZATION-strategies.md

Reference (2 frameworks):

  • REFERENCE-pm-mindset-strategy-execution.md
  • REFERENCE-additional-frameworks.md

All frameworks include:

  • Bilingual content (English/Vietnamese)
  • Real-world examples
  • Common mistakes
  • Keywords for further research (users research independently)
  • Internal navigation comments (for Claude only)

Data Storage System

Directory Structure

data/
├── items/              # Categorized PM items
│   ├── ideas/          # Product ideas (IDEA-001.md)
│   ├── incidents/      # Production incidents (INC-001.md)
│   ├── initiatives/    # Strategic initiatives
│   ├── ops-requests/   # Operational requests
│   ├── csat/           # Customer satisfaction issues
│   ├── integrations/   # Third-party needs
│   ├── funnel/         # Conversion optimization
│   ├── tech-debt/      # Technical debt
│   ├── competitive/    # Competitive responses
│   ├── compliance/     # Regulatory requirements
│   ├── insights/       # Data-driven insights
│   ├── research/       # User research
│   └── platform/       # Infrastructure needs
├── prds/              # Product Requirements Documents
├── roadmaps/          # Roadmap files
├── metrics/           # Analytics dashboards
├── retrospectives/    # Retrospective records
└── scores/            # Prioritization scores

File Format

All data files use YAML frontmatter:

---
id: "ITEM-001"
type: "idea"
status: "new"
date_created: "yyyy-mm-dd"
priority: "medium"
tags: []
related_items: []
---

Slash Commands

Available Commands (11 total)

Smart Assistant:

  • /assist [description] - AI-powered framework navigator that reads frameworks silently and provides synthesized guidance

Scenario-Specific:

  • /scenario:interview - PM interview preparation
  • /scenario:launch - Product launch & GTM planning
  • /scenario:churn - Churn analysis & retention
  • /scenario:prioritize - Feature prioritization
  • /scenario:feedback - Giving/receiving feedback
  • /scenario:pmf - Product-Market Fit validation
  • /scenario:chart - Data visualization guidance

Git Helpers:

  • /gh:commit - Analyze changes and create commit
  • /gh:init - Initialize git repository

Help & Guidance:

  • /prism-help - First-time user guide with real-world examples

Command Behavior

All slash commands MUST:

  1. Check conversation tracking first (PRISM-CONVOS/00-CONVO-LOGS.md)
  2. Read frameworks silently (from .claude/product-foundation/)
  3. Synthesize content (never tell users to read files)
  4. Provide keywords (for user self-research)
  5. Give actionable steps (extracted from frameworks)

Working with Frameworks

Framework Selection Guide

Use COMPETENCY-MAP.md and SCENARIO-INDEX.md to route user needs:

Common Scenarios:

  • "I need to prioritize features" → Read PRIORITIZATION-rice-prioritization.md, synthesize RICE methodology
  • "I'm launching a product" → Read LAUNCH-go-to-market-strategy.md, provide GTM checklist
  • "I don't know if we have PMF" → Read VALIDATION-product-market-fit.md, explain Sean Ellis test
  • "Users are churning" → Read METRICS-retention-churn-analysis.md, provide cohort analysis steps
  • "I'm interviewing someone" → Read PEOPLE-interviewing-framework.md, share question templates
  • "I need to influence stakeholders" → Read COMMUNICATION-stakeholder-management.md, give influence tactics
  • "Rural market access" → Read FINTECH-financial-inclusion.md, explain barriers & solutions
  • "Travel booking optimization" → Read TRAVELTECH-optimization-strategies.md, share conversion tactics

Framework Response Pattern

When helping users:

  1. Read the framework silently (they don't see this)
  2. Understand their scenario (restate their goal)
  3. Synthesize the framework (extract core methodology)
  4. Provide actionable steps (specific to their situation)
  5. Share keywords (for deeper self-research)
  6. Give pro tips (common mistakes from framework)

Example Template:

## 🎯 I Understand Your Scenario
[Restate what they're trying to accomplish]

## 📚 Framework Approach: [Name]
**Core methodology**: [Steps/formulas/models from framework]
**How to apply it**: [Specific guidance]
**Key concepts**: [2-3 principles]
**Common pitfalls**: [What to avoid]

## ⚡ Quick Action Plan
- [ ] [Specific action from framework]
- [ ] [Follow-up action]
- [ ] [Validation step]

**Time estimate**: [How long]

## 🎓 Key Takeaway
[One sentence insight]

## 💡 Pro Tips
[2-3 practical tips from framework]

## 🔍 Keywords for Further Research
- [Keyword 1]
- [Keyword 2]
- [Keyword 3]

Development Workflows

Common Tasks

No build/test commands - This is a PM knowledge repository, not a code project.

Key workflows:

  1. Using slash commands - Users invoke /assist or /scenario:* for PM guidance
  2. Managing conversations - Always check PRISM-CONVOS/00-CONVO-LOGS.md first
  3. Updating frameworks - Add new frameworks to .claude/product-foundation/ with bilingual content
  4. Creating slash commands - Add new .md files in .claude/commands/ following existing patterns

File Naming Conventions

  • Slash commands: Use hyphens, organize in folders (/scenario:feedback → .claude/commands/scenario/feedback.md)
  • Frameworks: Use PREFIX-name.md pattern (DISCOVERY-jobs-to-be-done.md)
  • Conversations: Use yyyy-mm-dd-topic-keywords.md pattern
  • Data files: Use PREFIX-NNN.md pattern (IDEA-001.md, INC-042.md)

Important Context

Project Identity

Name: PRISM (Lăng kính) Concept: Breaking complex PM problems into understandable components (like a prism with light) Version: PRISM v1.0 Previous names: Root v2.1 (multi-platform), Productify (original)

Project Philosophy

  • Capability over complexity: Frameworks are tools, not rules
  • Synthesis over signposting: Provide answers, not file paths
  • Keywords over links: Enable self-research, don't hand-hold
  • Context continuity: Track conversations to maintain coherence

Version History

  • Productify (original): Workflow-based PM assistant
  • Root v2.0: Multi-platform capability toolkit (5 AI platforms)
  • PRISM v1.0: Claude-only, production-ready, synthesis-focused

What Changed in PRISM v1.0

  1. ✅ Single platform (Claude Code only)
  2. ✅ Hidden frameworks (.claude/product-foundation/)
  3. ✅ Synthesis approach (Claude reads, users get insights)
  4. ✅ Keyword-based research (users Google independently)
  5. ✅ Streamlined structure (removed multi-platform overhead)
  6. ✅ Production ready (no development artifacts)

Quality Standards

Code Quality

  • Validate all file paths before operations
  • Use evidence-based claims (never assume completion)
  • Cross-reference documentation for consistency
  • Maintain bilingual content where applicable

Documentation Quality

  • Keep RULES.md and README.md synchronized
  • Update conversation logs with all interactions
  • Maintain accurate progress tracking
  • Document all framework additions

Framework Synthesis Quality

  • ✅ Read frameworks completely before responding
  • ✅ Extract actionable steps, not just theory
  • ✅ Provide real examples from frameworks
  • ✅ Share common mistakes from frameworks
  • ✅ Give 5-7 research keywords
  • ❌ Never say "read this file"
  • ❌ Never show file paths to users
  • ❌ Never recommend browsing .claude/product-foundation/

Getting Help

When in Doubt

  1. Check conversation history in PRISM-CONVOS/
  2. Reference RULES.md for protocols
  3. Read relevant frameworks in .claude/product-foundation/
  4. Synthesize and provide direct guidance

For Framework Navigation

  1. Check COMPETENCY-MAP.md for scenario mapping
  2. Check SCENARIO-INDEX.md for keyword routing
  3. Read matched frameworks silently
  4. Provide synthesized guidance with keywords

Remember: PRISM is about capability, not complexity. Frameworks are YOUR reference library - users get synthesized insights and research keywords, never file paths. Always maintain conversation continuity through PRISM-CONVOS/.