Skip to content

Latest commit

 

History

History
365 lines (260 loc) · 9.59 KB

projectbrief.md

File metadata and controls

365 lines (260 loc) · 9.59 KB

Claude Code Tresor - Project Brief

The Essential Claude Code Development Ecosystem

Version: 2.6.0 → 3.0 (In Progress) Last Updated: November 15, 2025

Project Vision

Mission: Become THE essential Claude Code companion that delivers 50-70% productivity improvements through an integrated ecosystem of skills, agents, commands, and standards.

Vision: Transform from a comprehensive agent library into a fully integrated development ecosystem where all components work together seamlessly to automate and enhance every aspect of software development.

Target Audience

Primary Users

1. Individual Developers

  • Solo developers using Claude Code
  • Need: Instant productivity boost, quality automation
  • Value: Professional-grade output, guided learning

2. Development Teams

  • Small to medium teams (5-50 developers)
  • Need: Consistent quality, enforced standards, collaboration
  • Value: Team alignment, knowledge sharing, efficiency

3. Enterprise Organizations

  • Large teams, compliance requirements
  • Need: Security, audit trails, custom standards
  • Value: Compliance, ROI tracking, governance

Secondary Users

4. Open Source Contributors

  • Community developers adding agents/skills
  • Need: Templates, guidelines, marketplace
  • Value: Recognition, adoption, impact

5. Technical Leaders

  • CTOs, architects, team leads
  • Need: Strategic tooling, metrics, decision support
  • Value: Team productivity, quality metrics, cost savings

Core Components

1. Subagents (141 total)

Purpose: Specialized AI assistants for specific domains

Categories (10):

  • Core (8): Essential production-ready agents
  • Engineering (54): Software development specialists
  • Design (7): UI/UX and visual design
  • Marketing (11): Content and growth
  • Product (9): Product management
  • Leadership (14): Finance and strategy
  • Operations (6): Business operations
  • Research (7): Market intelligence
  • AI/Automation (9): AI/ML specialists
  • Account/CS (8): Customer success

Quality: 9.7/10 (exceptional)

2. Skills (8 → 20 target)

Purpose: Autonomous background helpers that auto-detect issues

Current:

  • Development: code-reviewer, test-generator, git-commit-helper
  • Security: security-auditor, secret-scanner, dependency-auditor
  • Documentation: api-documenter, readme-updater

Planned:

  • Performance: performance-monitor, bundle-analyzer
  • Quality: accessibility-checker, react-best-practices, python-style-checker
  • DevOps: docker-validator, env-validator, migration-safety

Integration: Skills detect → Recommend specific agents → One-click invocation

3. Commands (4 → 15 target)

Purpose: Workflow orchestration and automation

Current:

  • /scaffold - Project/component generation
  • /review - Code review automation
  • /test-gen - Test generation
  • /docs-gen - Documentation generation

Planned:

  • /audit - Security audit workflow
  • /deploy - Deployment automation
  • /migrate - Database migration workflow
  • /optimize - Performance optimization
  • /refactor - Refactoring workflow
  • /discover-agent - Agent discovery wizard
  • /enforce-standard - Standards validation
  • /incident - Production incident response
  • /analyze - Codebase analysis
  • /release - Release preparation

Integration: Commands orchestrate agents + activate skills + enforce standards

4. Standards (12 → 30 target)

Purpose: Development standards and best practices

Current:

  • Style guides: JavaScript, TypeScript, Python, React, CSS
  • Workflows: Git conventional commits
  • Templates: PR, Issue, README, API docs

Planned:

  • Languages: Rust, Swift, Kotlin, SQL, YAML
  • Frameworks: Next.js, Vue 3, Django, Spring Boot
  • Architecture: Microservices, Event-driven, API design
  • Processes: ADR, Code review, Testing, Security baseline

Vision: Living standards enforced by agents with auto-fix

5. Prompts (7 → 40 target)

Purpose: Guided development templates

Current:

  • Architecture, code generation, debugging, best practices

Planned:

  • AI/ML: Pipeline design, RAG systems, LLM applications
  • Cloud: AWS, Kubernetes, Terraform, Serverless
  • Mobile: React Native, Flutter, iOS, Android
  • Domains: Fintech, Healthcare, E-commerce, IoT

Integration: Prompts include recommended agent workflows

Architecture & Tech Stack

Repository Structure

claude-code-tresor/
├── agents/ (8 core agents)
├── subagents/ (133 specialized agents)
├── skills/ (8 → 20 autonomous helpers)
├── commands/ (4 → 15 orchestration workflows)
├── standards/ (12 → 30 enforced standards)
├── prompts/ (7 → 40 guided templates)
├── examples/ (12 → 50 workflow examples)
├── scripts/ (installation utilities)
├── indexes/ (NEW - machine-readable catalogs)
├── docs/ (consolidated documentation)
└── Memory bank (projectbrief, productContext, activeContext)

Technology Stack

Core Technologies:

  • Markdown (documentation, agent definitions)
  • YAML (frontmatter, configuration, workflows)
  • JSON (indexes, metadata)
  • Bash (installation scripts, automation)
  • Python (validation, analysis, tooling)

Claude Code Integration:

  • Sub-agents (Markdown with YAML frontmatter)
  • Skills (SKILL.md with trigger patterns)
  • Slash commands (Command frontmatter)

Future:

  • CLI tool (Node.js or Python)
  • Web dashboard (React + Tailwind)
  • Analytics (usage tracking)

Component Taxonomy

Skills

Type: Autonomous background helpers Activation: Automatic on file changes Tools: Limited (Read, Grep, Bash, Edit) Purpose: Quick detection, 3-5 suggestions

Categories:

  1. Development: Code quality, testing, git
  2. Security: Vulnerability detection, secrets, dependencies
  3. Documentation: API docs, README maintenance
  4. Performance: Bottleneck detection, optimization
  5. DevOps: Docker, environment, migrations

Subagents

Type: Specialized expert assistants Activation: Manual (@agent-name) Tools: Full access (Read, Write, Edit, Bash, Task, etc.) Purpose: Comprehensive analysis, detailed recommendations

Organization:

  • 10 team categories (color-coded)
  • 40+ functional subcategories
  • Clear specialization and expertise

Commands

Type: Workflow orchestration Activation: Manual (/command-name) Tools: Task (to invoke agents) Purpose: Multi-agent workflows, automation

Pattern: Command → Agents → Skills → Output

Standards

Type: Development best practices Activation: Manual enforcement (/enforce-standard) Purpose: Quality gates, consistency, compliance

Evolution: Static docs → Agent-enforced → Auto-fix

Prompts

Type: Guided development templates Activation: Manual (copy/paste or use in Claude) Purpose: Structured problem-solving, best practices

Enhancement: Include agent workflow recommendations

Success Definition

Adoption Metrics

Week 4: 20% of users adopt new features Week 8: 50% regular usage of discovery/commands Week 12: 70% team standardization Week 16: Community contributions, marketplace launch

Productivity Metrics

Immediate: 40% faster with discovery + audit + deploy Mid-term: 50% faster feature development Long-term: 70% overall productivity improvement

Quality Metrics

Standards Compliance: 90%+ Test Coverage: 95%+ on new code Security: 60% fewer vulnerabilities Performance: 99% uptime (config safety)

Business Metrics

GitHub Stars: 1000+ (from ~100) Enterprise Adoption: 5-10 teams Community Agents: 20+ contributed Industry Recognition: Featured in dev publications

Strategic Priorities

P0 - Critical Path:

  1. Fix installer (blocks everything)
  2. Create memory bank (ensures continuity)
  3. Generate indexes (enables tooling)
  4. /discover-agent (solves overwhelming choice)

P1 - High Value: 5. Essential skills (performance, accessibility, docker, env) 6. Critical commands (/audit, /deploy, /migrate) 7. Standard enforcement 8. Discovery system

P2 - Ecosystem Integration: 9. Skill-agent coordination 10. Command-skill activation 11. Prompt-agent workflows 12. Example execution

P3 - Advanced Features: 13. Workflow composer 14. Command chaining 15. Marketplace 16. Analytics

Risk Management

Technical Risks:

  • Complexity → Mitigate: Incremental development, continuous validation
  • Breaking changes → Mitigate: Backward compatibility, migration guides
  • Performance → Mitigate: Profiling, optimization, caching

Adoption Risks:

  • Overwhelming features → Mitigate: Progressive disclosure, onboarding
  • Learning curve → Mitigate: Examples, videos, quick-starts
  • Discovery → Mitigate: Wizard, recommendations, documentation

Community Risks:

  • Low engagement → Mitigate: Marketing, showcases, responsiveness
  • Quality variations → Mitigate: Marketplace reviews, validation
  • Support burden → Mitigate: Documentation, FAQs, community

Next Immediate Actions

This Week (Week 1):

Day 1:

  • Create memory bank files (4 hours)
  • Generate machine-readable indexes (4 hours)

Day 2:

  • Fix installer metadata (6 hours)
  • Test installation

Day 3:

  • Audit and fix documentation (3 hours)
  • Start /discover-agent command (5 hours)

Day 4:

  • Complete /discover-agent (3 hours)
  • Create /audit command (5 hours)

Day 5:

  • Start performance-monitor skill (6 hours)
  • Testing and validation (2 hours)

Week 1 Total: 38 hours (feasible over 5 days)

Created: November 15, 2025 Owner: Alireza Rezvani Contributors: Claude (Anthropic AI) Status: Ready for Phase 1 Execution