telo3-ai 5.0.7 → 6.2.0

package/README.md

@@ -1,818 +1,59 @@
 # Telo3
 
-**Telo3 is an AI development framework and project-context system for coding agents. Telo3 provides persistent memory, quality frameworks, and systematic engineering practices that transform AI coding assistants (Claude, ChatGPT, GitHub Copilot, Cursor) into reliable development partners.**
+A chat-first skill for less AI slop, less wasted context, and continuity between coding sessions. Install in your project, say **telo3** once, then work normally. No mandatory CLI use after installation.
 
-[![MIT License](https://img.shields.io/badge/License-MIT-green.svg)](https://opensource.org/licenses/MIT)
-[![GitHub stars](https://img.shields.io/github/stars/jeet1511/telo3?style=social)](https://github.com/jeet1511/telo3/stargazers)
-[![GitHub forks](https://img.shields.io/github/forks/jeet1511/telo3?style=social)](https://github.com/jeet1511/telo3/network/members)
-[![PRs Welcome](https://img.shields.io/badge/PRs-welcome-brightgreen.svg)](https://github.com/jeet1511/telo3/pulls)
-[![Author](https://img.shields.io/badge/Author-Jeet-blue?style=flat&logo=github)](https://github.com/jeet1511)
+## Install
 
-> **Created by [Jeet (@jeet1511)](https://github.com/jeet1511)** · [Repository](https://github.com/jeet1511/telo3) · [Issues](https://github.com/jeet1511/telo3/issues) · [Contributing](CONTRIBUTING.md)
-
----
-
-## What is Telo3?
-
-Telo3 is a comprehensive AI development framework that provides project-context management, quality assurance frameworks, and systematic engineering workflows for AI coding agents. The Telo3 system prevents AI hallucination, architectural drift, and inconsistent code quality through a 6-document context architecture and 11 integrated quality frameworks.
-
-### Why Telo3?
-
-Traditional AI coding assistants suffer from context loss, inconsistent quality, and "vibe coding" (generic, aesthetically-driven code without systematic engineering). Telo3 solves these problems through:
-
-**Problems Telo3 Prevents:**
-
-- ❌ **AI Vibe Coding** — Generic aesthetics without engineering discipline
-- ❌ **Context Loss** — Forgotten decisions and repeated explanations
-- ❌ **Token Inefficiency** — Wasteful token usage and slow iteration
-- ❌ **Security Vulnerabilities** — Unvalidated input and exposed credentials
-- ❌ **Privacy Violations** — Undocumented tracking and compliance gaps
-- ❌ **Architectural Drift** — Inconsistent patterns and random technology choices
-- ❌ **Design Inconsistency** — Copy-paste interfaces without design systems
-- ❌ **Quality Gaps** — Missing accessibility, SEO, and performance standards
-
-**How Telo3 Solves These:**
-
-**How Telo3 Solves These:**
-
-- ✅ **6-Document Project Context** — prd.md, architecture.md, rules.md, design.md, tasks.md, memory.md provide persistent AI memory
-- ✅ **11 Quality Frameworks** — Security, privacy, accessibility (WCAG), SEO, performance (Core Web Vitals), legal compliance (GDPR, COPPA, CAN-SPAM, DMCA)
-- ✅ **Token Optimization** — 80% reduction through selective context loading and "caveman mode" communication
-- ✅ **Smart Delegation** — Automatic detection and use of specialized skills and MCP servers
-- ✅ **Anti-Vibe-Coding Framework** — Systematic design system enforcement
-- ✅ **Framework Agnostic** — Works with React, Next.js, Vue, Python, Go, Rust, and any technology stack
-- ✅ **Smart Onboarding** — AI-driven project discovery and automated setup for empty projects
-
----
-
-## Telo3 Architecture
-
-## Telo3 Architecture
-
-The Telo3 system consists of three core layers:
-
-### 1. Project Context Layer (6-Document System)
-
-Telo3 maintains project state through six interconnected markdown documents:
-
-| Document | Purpose | Telo3 Function |
-|----------|---------|----------------|
-| **prd.md** | Product requirements | What the project does and why |
-| **architecture.md** | Technical architecture | System structure, technology choices, patterns |
-| **rules.md** | Coding standards | Language-specific conventions, best practices |
-| **design.md** | Design system | UI/UX standards, component library, brand guidelines |
-| **tasks.md** | Implementation roadmap | Current sprint, backlog, completed work |
-| **memory.md** | Decision log | Recent changes, active context, session continuity |
-
-### 2. Quality Framework Layer (11 Dimensions)
-
-Telo3 quality frameworks ensure production-ready code:
-
-| Framework | Standards | Telo3 Reference |
-|-----------|-----------|-----------------|
-| **Security** | OWASP, input validation, authentication | [security-quality.md](references/security-quality.md) |
-| **Privacy** | GDPR, COPPA, data protection | [privacy-compliance.md](references/privacy-compliance.md) |
-| **Legal** | CAN-SPAM, DMCA, terms compliance | [legal-compliance.md](references/legal-compliance.md) |
-| **Accessibility** | WCAG 2.1 AA, keyboard navigation, screen readers | [accessibility-quality.md](references/accessibility-quality.md) |
-| **SEO** | Meta tags, semantic HTML, Core Web Vitals | [seo-quality.md](references/seo-quality.md) |
-| **Performance** | LCP, INP, CLS optimization | [performance-quality.md](references/performance-quality.md) |
-| **UI Quality** | Design system consistency, anti-vibe-coding | [ui-quality.md](references/ui-quality.md) |
-| **Token Efficiency** | Selective loading, caveman mode | [token-optimization.md](references/token-optimization.md) |
-| **Smart Delegation** | MCP integration, skill discovery | [smart-skill-delegation.md](references/smart-skill-delegation.md) |
-| **Documentation** | Systematic specifications | [document-specification.md](references/document-specification.md) |
-| **Workflow** | AI behavior patterns | [workflow.md](references/workflow.md) |
-
-### 3. Intelligence Layer (Smart Systems)
-
-- **Smart Onboarding** — Automated project discovery and planning ([ONBOARDING-SKILL.md](ONBOARDING-SKILL.md))
-- **Context Intelligence** — Selective document loading based on task
-- **Skill Delegation** — Automatic detection of specialized tools
-- **Token Optimization** — 80% reduction through intelligent caching
-
----
-
-## Features
-
-## Features
-
-### Core Telo3 Capabilities
-
-- **Persistent Project Memory** — Six markdown documents (prd, architecture, rules, design, tasks, memory) maintain complete project context across AI sessions
-- **Smart Onboarding** — Automated project discovery through conversational AI for empty codebases ([Quick Start Guide](QUICK-START-ONBOARDING.md))
-- **Token Optimization** — 80% reduction through selective context loading and caveman mode communication
-- **Quality Frameworks** — 11 production-ready frameworks covering security, privacy, accessibility, SEO, performance
-- **Smart Skill Delegation** — Automatic detection and integration of specialized MCP servers and AI skills
-- **Framework Agnostic** — Compatible with React, Vue, Next.js, Python, Go, Rust, and any technology stack
-- **Universal AI Compatibility** — Works with Claude, ChatGPT, GitHub Copilot, Cursor, Kiro, and any AI coding assistant
-
-### Telo3 Quality Dimensions
-
-Telo3 enforces production standards across 11 quality dimensions:
-
-1. **Security** — OWASP standards, input validation, authentication patterns
-2. **Privacy** — GDPR, COPPA, CCPA compliance guidance
-3. **Legal** — CAN-SPAM, DMCA, terms of service frameworks
-4. **Accessibility** — WCAG 2.1 Level AA compliance
-5. **SEO** — Semantic HTML, meta tags, Core Web Vitals
-6. **Performance** — LCP, INP, CLS optimization strategies
-7. **UI Quality** — Design system consistency, anti-vibe-coding principles
-8. **Token Efficiency** — Context caching and selective loading
-9. **Smart Delegation** — MCP and skill integration intelligence
-10. **Documentation** — Systematic specification standards
-11. **Workflow** — AI behavior and decision patterns
-
-### Efficiency Modes
-
-Telo3 supports multiple communication modes optimized for different use cases:
-
-| Mode | Description | Token Savings | Use Case |
-|------|-------------|---------------|----------|
-| **Normal** | Clear, concise documentation | 50% | Standard development |
-| **Caveman** | Ultra-compressed communication | 80% | Rapid iteration |
-| **Selective** | Context-aware document loading | Varies | Large projects |
-| **Smart Delegation** | Automatic tool selection | Varies | Specialized tasks |
-
----
-
-## Installation
-
-Telo3 is available via npm, skillpm, and GitHub.
-
-### Install via npm
-
-```bash
-# Install globally (recommended)
-npm install -g telo3-ai
-
-# Or install in your project
-npm install telo3-ai
-
-# Or use npx (no installation needed)
-npx telo3-ai init
-```
-
-**Package:** https://www.npmjs.com/package/telo3-ai
-
-### Install via skillpm
+In your project, run one short command:
 
 ```bash
-# Install as AI skill
-npx skillpm install telo3-ai
-
-# List installed skills
-npx skillpm list
-```
-
-**Registry:** https://skillpm.dev
-
-### Or Clone from GitHub
-
-```bash
-# Clone Telo3 to your home directory
-git clone https://github.com/Jeet1511/telo3.git ~/telo3
-
-# Or clone into your project directory
-git clone https://github.com/Jeet1511/telo3.git ./telo3
-```
-
-### Or Add as Git Submodule
-
-```bash
-# Add Telo3 to an existing project
-cd your-project
-git submodule add https://github.com/Jeet1511/telo3.git telo3
-```
-
----
-
-## Usage
-
-### Initialize Telo3 Project Context
-
-**If installed via npm:**
-```bash
-cd your-project
-telo3 init
-```
-
-**If cloned from GitHub:**
-```bash
-cd your-project
-~/telo3/scripts/init-project-context.sh
-```
-
-**Manual setup:**
-```bash
-# Copy Telo3 templates to your project
-cp -r ~/telo3/templates/* ./project-context/
-```
-
-### Activate Telo3 with Your AI
-
-**For new/empty projects:**
-```
-Use Telo3 onboarding - this is a new project
-```
-
-**For existing projects:**
-
-**For existing projects:**
-
-**Kiro IDE:**
-```
-Use Telo3 from ~/telo3/ - Read AI-CONTEXT.md first
-```
-
-**Cursor:**
-```
-@telo3/AI-CONTEXT.md Use Telo3 systematic approach
-```
-
-**GitHub Copilot:**
-```
-#file:telo3/AI-CONTEXT.md Follow Telo3 project context system
-```
-
-**Claude / ChatGPT:**
-```
-Use Telo3 framework from ~/telo3/ - Read AI-CONTEXT.md for quick reference
-```
-
----
-
-## Telo3 Skills
-
-Telo3 provides specialized skills for different development scenarios:
-
-- **[SKILL.md](SKILL.md)** — Complete Telo3 system documentation and AI instructions
-- **[ONBOARDING-SKILL.md](ONBOARDING-SKILL.md)** — Smart project discovery and automated setup for empty projects
-- **[AI-CONTEXT.md](AI-CONTEXT.md)** — 30-second quick reference for AI assistants
-
-## Telo3 Project Context
-
-The Telo3 project-context system consists of six core documents that maintain persistent state:
-
-1. **[prd.md](templates/prd.md)** — Product requirements and project goals
-2. **[architecture.md](templates/architecture.md)** — System architecture and technical decisions
-3. **[rules.md](templates/rules.md)** — Coding standards and best practices
-4. **[design.md](templates/design.md)** — Design system and UI/UX guidelines
-5. **[tasks.md](templates/tasks.md)** — Implementation roadmap and task tracking
-6. **[memory.md](templates/memory.md)** — Recent decisions and active context
-
-## Supported AI Coding Tools
-
-Telo3 works with any AI coding assistant, including:
-
-- **Claude** (Anthropic) — Excellent systematic reasoning
-- **ChatGPT** (OpenAI) — Fast prototyping and code generation
-- **GitHub Copilot** — IDE-integrated suggestions
-- **Cursor** — AI-first code editor
-- **Kiro** — AI development environment
-- Any other AI assistant that can read markdown documentation
-
-## Documentation
-
-### Core Documentation
-- **[README.md](README.md)** — This file, Telo3 overview
-- **[SKILL.md](SKILL.md)** — Complete Telo3 instructions for AI
-- **[AI-CONTEXT.md](AI-CONTEXT.md)** — Quick reference guide
-- **[ONBOARDING-SKILL.md](ONBOARDING-SKILL.md)** — Smart onboarding system
-- **[QUICK-START-ONBOARDING.md](QUICK-START-ONBOARDING.md)** — Onboarding quick start
-
-### Quality Framework References
-- [Token Optimization](references/token-optimization.md) — 80% token savings through selective loading
-- [Smart Skill Delegation](references/smart-skill-delegation.md) — Automatic MCP and skill integration
-- [UI Quality](references/ui-quality.md) — Anti-vibe-coding design principles
-- [Security](references/security-quality.md) — OWASP standards and secure coding
-- [Privacy](references/privacy-compliance.md) — GDPR, COPPA, CCPA guidance
-- [Legal](references/legal-compliance.md) — CAN-SPAM, DMCA, compliance
-- [SEO](references/seo-quality.md) — Semantic HTML, Core Web Vitals
-- [Accessibility](references/accessibility-quality.md) — WCAG 2.1 AA standards
-- [Performance](references/performance-quality.md) — LCP, INP, CLS optimization
-- [Document Specification](references/document-specification.md) — Documentation standards
-- [Workflow](references/workflow.md) — AI behavior patterns
-- [AI Behavior](references/ai-behavior.md) — Decision-making frameworks
-- [Project Audit](references/project-audit.md) — Quality assessment checklist
-
----
-
-## Author
-
-**Telo3** is created and maintained by **[Jeet (@jeet1511)](https://github.com/jeet1511)**.
-
-- GitHub: [@jeet1511](https://github.com/jeet1511)
-- Repository: [github.com/Jeet1511/telo3](https://github.com/Jeet1511/telo3)
-- Issues: [github.com/Jeet1511/telo3/issues](https://github.com/Jeet1511/telo3/issues)
-
----
-
-## License
-
-Telo3 is open source software licensed under the **[MIT License](LICENSE)**.
-
-You are free to use, modify, and distribute Telo3 for any purpose, including commercial projects.
-
----
-| **design.md** | HOW it looks - Visual language, components | When design patterns change |
-| **tasks.md** | WHAT remains - Implementation backlog | Weekly or after work |
-| **memory.md** | CURRENT STATE - Recent work, decisions | After important changes |
-
-### Why This Works
-
-**Before Telo3:**
-- AI forgets project decisions
-- Repeats same questions
-- Ignores established patterns
-- Makes random choices
-- Wasts tokens reading everything
-
-**With Telo3:**
-- Persistent project memory
-- Understands requirements
-- Follows established patterns
-- Makes informed decisions
-- Efficient token usage
-
----
-
-## 🎨 Anti-Vibe-Coding System
-
-Telo3 prevents generic AI aesthetics:
-
-### What It Catches
-
-❌ Harsh/arbitrary gradients  
-❌ Generic purple/black "AI" palettes  
-❌ Excessive shadows and effects  
-❌ Inconsistent design  
-❌ Generic SaaS layouts  
-❌ Decorative UI without purpose  
-❌ Copy-paste components  
-❌ Fake statistics/testimonials  
-
-### What It Enforces
-
-✅ Follow actual design system  
-✅ Reuse existing components  
-✅ Consistent visual language  
-✅ Purpose-driven UI  
-✅ Brand-appropriate aesthetics  
-
----
-
-## 🔒 Security & Privacy Built-In
-
-### Security Checks
-
-- ✅ Input validation (all user input)
-- ✅ Authorization (every API endpoint)
-- ✅ Secret protection (never committed)
-- ✅ SQL injection prevention
-- ✅ XSS prevention
-- ✅ CSRF protection
-- ✅ Security headers
-
-### Privacy Safeguards
-
-- ✅ Data minimization
-- ✅ Third-party audit
-- ✅ Cookie transparency
-- ✅ Self-hosted fonts (no Google Fonts by default)
-- ✅ Session replay safeguards
-- ✅ Children's privacy (COPPA awareness)
-- ✅ Data deletion support
-
-### Legal Compliance Awareness
-
-> **Important:** Telo3 provides engineering guidance, not legal advice.
-
-- ✅ Age restrictions (COPPA for U.S.)
-- ✅ Marketing emails (CAN-SPAM, GDPR)
-- ✅ Subscriptions (disclosure requirements)
-- ✅ User content (DMCA considerations)
-- ✅ Recommends legal counsel verification
-
----
-
-## 🚄 Token Optimization
-
-### How Telo3 Saves 80% Tokens
-
-**Traditional AI:**
-1. Reads entire codebase (2000 tokens)
-2. Reads all documentation (1500 tokens)
-3. Runs unnecessary tests (500 tokens)
-4. Verbose explanations (1000 tokens)
-**Total:** ~5000 tokens per task
-
-**With Telo3:**
-1. Reads memory.md only (100 tokens)
-2. Reads affected file (200 tokens)
-3. Skips unnecessary tests (0 tokens)
-4. Caveman mode response (50 tokens)
-**Total:** ~350 tokens per task
-
-**Savings:** 93% on this example!
-
-### Caveman Mode
-
-```
-User: "Use caveman mode"
-AI: "Caveman mode ON. Short responses. Ready."
-
-User: "Add login button"
-AI: "Done.
-
-Changed:
-- LoginButton.tsx (new)
-- auth.ts (imported)
-
-Follows design.md button styles.
-Works."
-```
-
----
-
-## 🔧 Smart Skill Delegation
-
-Telo3 automatically checks for better tools:
-
-### Before Implementing
-
-**Smart AI behavior:**
-- Checks for Figma power (design implementation)
-- Checks for context7 MCP (library docs)
-- Checks for specialized skills
-- Recommends MCP configuration if beneficial
-- Uses best tool for each job
-
-**Example:**
-```
-Task: Implement Figma design
-
-Telo3: "Figma power available. Using for accurate extraction.
-Faster + more accurate than manual. Proceeding..."
-```
-
----
-
-## 📦 What's Included
-
+npm i telo3-ai
 ```
-telo3/
-├── AI-CONTEXT.md               # 🔥 READ THIS FIRST
-├── SKILL.md                    # Complete AI instructions
-├── README.md                   # This file
-├── SKILL-METADATA.json         # AI discovery metadata
-├── LICENSE                     # MIT License
-├── CHANGELOG.md                # Version history
-│
-├── templates/ (6 files)        # Core document templates
-│   ├── prd.md
-│   ├── architecture.md
-│   ├── rules.md
-│   ├── design.md
-│   ├── tasks.md
-│   └── memory.md
-│
-├── references/ (13 files)      # Quality frameworks
-│   ├── token-optimization.md           # 🔥 80% savings
-│   ├── smart-skill-delegation.md       # 🔥 Use better tools
-│   ├── ui-quality.md                   # Anti-vibe-coding
-│   ├── seo-quality.md                  # SEO framework
-│   ├── accessibility-quality.md        # WCAG compliance
-│   ├── performance-quality.md          # Core Web Vitals
-│   ├── security-quality.md             # Security audit
-│   ├── privacy-compliance.md           # Privacy guidelines
-│   ├── legal-compliance.md             # Legal awareness
-│   ├── document-specification.md
-│   ├── workflow.md
-│   ├── ai-behavior.md
-│   └── project-audit.md
-│
-├── examples/example-project/ (6 files)  # Complete example
-│
-└── scripts/ (2 files)           # Automation
-    ├── init-project-context.sh
-    └── validate-project-context.sh
-```
-
-**Total:** 31 files, ~280KB of systematic engineering knowledge
-
----
-
-## 🌟 Real-World Benefits
-
-### For AI Agents
-- ✅ Persistent memory across sessions
-- ✅ Clear engineering standards
-- ✅ Decision-making framework
-- ✅ Quality guardrails
-- ✅ Token efficiency
-
-### For Developers
-- ✅ Consistent AI behavior
-- ✅ Documented decisions
-- ✅ Quality assurance
-- ✅ Faster onboarding
-- ✅ Reduced drift
-
-### For Projects
-- ✅ Design consistency
-- ✅ Security by design
-- ✅ Accessibility compliance
-- ✅ Performance optimization
-- ✅ Legal awareness
-
----
-
-## 🎯 Use Cases
-
-### Perfect For
-
-| Project Type | Why Telo3 Helps |
-|--------------|-----------------|
-| **Startups** | Systematic from day 1, scales properly |
-| **SaaS Products** | Security, privacy, compliance built-in |
-| **E-commerce** | Payment security, subscriptions, SEO |
-| **Portfolios** | SEO, accessibility, design consistency |
-| **Enterprise** | Documentation, architecture, quality |
-| **Open Source** | Clear contribution guidelines, standards |
-
-### Works With
-
-**ANY AI coding assistant:**
-- ✅ **Kiro IDE**
-- ✅ **Cursor** 
-- ✅ **GitHub Copilot** (Chat & Workspace)
-- ✅ **Claude** (Desktop, API, Chat)
-- ✅ **ChatGPT** (Plus, API)
-- ✅ **Gemini** (Code Assist, Chat)
-- ✅ **Codeium**
-- ✅ **Tabnine**
-- ✅ **Amazon CodeWhisperer**
-- ✅ **VS Code with any AI extension**
-- ✅ **JetBrains with AI Assistant**
-- ✅ **Any IDE + Any AI**
-
-**How it works universally:**
-1. Clone Telo3 to any folder
-2. Initialize project context anywhere
-3. Reference files from your AI tool
-4. AI follows systematic approach
-
-**No platform lock-in. Works everywhere.**
-
-### Framework Support
-
-- ✅ React, Next.js, Vue, Svelte, Angular
-- ✅ Node.js, Python, Go, Rust, PHP
-- ✅ Any tech stack (framework-agnostic)
-
----
-
-## 📖 Documentation
-
-### Quick References
-- [AI-CONTEXT.md](AI-CONTEXT.md) - 30-second quick start
-- [SKILL.md](SKILL.md) - Complete AI instructions
-- [CHANGELOG.md](CHANGELOG.md) - Version history
-
-### Quality Frameworks
-- [Token Optimization](references/token-optimization.md) - Save 80% tokens
-- [Smart Delegation](references/smart-skill-delegation.md) - Use better tools
-- [UI Quality](references/ui-quality.md) - Anti-vibe-coding
-- [Security](references/security-quality.md) - Security audit
-- [Privacy](references/privacy-compliance.md) - Privacy guidelines
-- [Legal](references/legal-compliance.md) - Compliance awareness
-- [SEO](references/seo-quality.md) - SEO optimization
-- [Accessibility](references/accessibility-quality.md) - WCAG standards
-- [Performance](references/performance-quality.md) - Core Web Vitals
 
-### Full Documentation
-- [Document Specification](references/document-specification.md)
-- [AI Workflow](references/workflow.md)
-- [AI Behavior](references/ai-behavior.md)
-- [Project Audit](references/project-audit.md)
+That's it. The install adds the skill to your project and activates the always-on contract automatically (additive, idempotent, never overwrites your files; skips global installs). Then say **telo3** once in your AI chat.
 
----
+Prefer npx, or a specific builder? `npx telo3-ai` does the same; `npx telo3-ai install --agent qoder` targets a host (`claude`, `cursor`, `gemini`, `copilot`, `codex`, `agents`). Requires Node.js 20+.
 
-## 🤝 Contributing
+The skill is copied into your project, independent of npm's cache. It detects existing Qoder, Claude, Gemini or Cursor folders; otherwise uses the shared `.agents/skills/telo3` directory. Refresh the builder's skill list if it caches discovery. Future sessions resume from project instructions and the short memory summary.
 
-Contributions welcome! Telo3 is open source and community-driven.
+No terminal at all? Copy this package into the project's skill directory as `telo3/` (with `SKILL.md`, references, templates and runtime assets), then say **telo3**. Plain web chat without project-file access cannot persist activation by itself.
 
-### How to Contribute
+## Small core, on-demand depth
 
-1. Fork the repository
-2. Create feature branch (`git checkout -b feature/amazing`)
-3. Commit changes (`git commit -m 'Add amazing feature'`)
-4. Push to branch (`git push origin feature/amazing`)
-5. Open Pull Request
+One compact contract in `AGENTS.md`; other host files contain small pointers, not duplicate full instructions. Domain references load only when needed. The skill is opt-in, not a prompt that hijacks every unrelated project.
 
-### Areas for Contribution
+Six context documents have separate roles: requirements (`prd.md`), structure (`architecture.md`), engineering (`rules.md`), UI/UX (`design.md`), work (`tasks.md`), handoff (`memory.md`). Reuse existing docs; create only missing context appropriate to substantial work. Memory is a rolling summary, not a transcript or personal memory store.
 
-- Additional quality frameworks
-- Language-specific templates
-- Framework-specific examples
-- Translation (i18n)
-- Documentation improvements
-- Bug fixes
+Keep security/privacy, purposeful design, accessibility and performance checks relevant to the actual change. The 14 references retain detailed guidance without loading every audit every turn. Host permissions and current user scope remain authoritative; the skill cannot override them or guarantee model compliance.
 
----
+## Optional commands
 
-## 📜 License
+| Command | Result |
+|---|---|
+| `install [dir] --agent <host>` | Durable skill installation; preserves user edits |
+| `activate [dir] [--all]` | Persistent core + host pointers; repeat-safe |
+| `deactivate [dir]` | Remove only Telo3 instruction blocks; keep custom content |
+| `init [dir] --context-dir <relative-path>` | Six skeletons; never overwrite existing docs |
+| `validate [dir] [--json] [--strict]` | Sections, placeholders, likely secrets and bounded stale scan |
+| `status [dir]` | Context health |
+| `context <task> [dir] --measure` | Character-based estimate of context-loading reduction |
+| `version` | Package version |
 
-MIT License - see [LICENSE](LICENSE) file.
+Token comparison uses selected docs + task references versus **all six docs + the same references**. It is not an actual model/API token counter. Missing selected docs are reported. Detected credentials are redacted from context output; this is heuristic protection, not complete data-loss prevention.
 
-Copyright (c) 2024 **[Jeet](https://github.com/jeet1511)**
+Installation refuses unmanaged directories or modified installer-owned files. Updates are staged with rollback; backups survive if recovery itself fails. Do not edit skill files concurrently during an update.
 
----
+## Optional Claude Code hook
 
-## 👨‍💻 Author
+`hooks/session-start.example.json` shows opt-in registration. Adjust the command to the installed skill's location. The hook returns only the core from an already-activated project's `AGENTS.md`; it does not silently activate unrelated projects or load all memory. No hooks/settings are installed without user authorization. Host project instructions remain the primary persistence mechanism.
 
-**Jeet**  
-GitHub: [@jeet1511](https://github.com/jeet1511)  
-Created: December 2024
-
----
-
-## 🌟 Show Your Support
-
-If Telo3 helped you, please:
-
-⭐ **Star this repo** on GitHub  
-🍴 **Fork** and contribute  
-📢 **Share** with your team  
-💬 **Report issues** or suggest features  
-☕ **[Buy me a coffee](https://buymeacoffee.com/jeet1511)**
-
----
-
-## 📊 Stats & Metrics
-
-- **Token Savings:** 80% average
-- **Speed Increase:** 5x faster tasks
-- **Quality Dimensions:** 11 frameworks
-- **Supported Platforms:** 7+ AI tools
-- **Framework Support:** All major frameworks
-- **License:** MIT (free forever)
-- **Files:** 31 comprehensive files
-- **Community:** Growing daily
-
----
-
-## 🔮 Roadmap
-
-### v1.1 (Coming Soon)
-- [ ] Visual documentation generator
-- [ ] Integration with more MCP servers
-- [ ] Multi-language support
-- [ ] Video tutorials
-- [ ] Community skill registry
-
-### v1.2 (Future)
-- [ ] Auto-audit scheduling
-- [ ] AI-powered context optimization
-- [ ] Team collaboration features
-- [ ] Cloud sync option
-
----
-
-## 💬 Community
-
-- **Discord:** [Join our community](https://discord.gg/telo3)
-- **GitHub Discussions:** [Ask questions](https://github.com/jeet1511/telo3/discussions)
-- **Issues:** [Report bugs](https://github.com/jeet1511/telo3/issues)
-- **Twitter:** [@jeet1511](https://twitter.com/jeet1511)
-
----
-
-## 🏆 Why Telo3 is Different
-
-### vs Traditional Prompt Engineering
-- ❌ Prompts: Lost after session, inconsistent
-- ✅ Telo3: Persistent memory, systematic
-
-### vs Generic AI Coding
-- ❌ Generic: Vibe coding, random decisions
-- ✅ Telo3: Follows system, informed decisions
-
-### vs Manual Documentation
-- ❌ Manual: Gets stale, ignored by AI
-- ✅ Telo3: Always current, AI understands
-
-### vs Other Skills
-- ❌ Others: Single-purpose, limited scope
-- ✅ Telo3: Comprehensive, 11 quality dimensions
-
----
-
-## 🎬 Quick Demo
+## Release verification (source repository)
 
 ```bash
-# 1. Clone Telo3
-git clone https://github.com/jeet1511/telo3.git ~/telo3
-
-# 2. Initialize your project
-cd my-project
-~/telo3/scripts/init-project-context.sh
-
-# 3. Start your AI coding session
-kiro
-
-# 4. Tell AI
-"Use Telo3. Read AI-CONTEXT.md. Enable caveman mode."
-
-# 5. Watch the magic ✨
-# AI now codes systematically with 80% fewer tokens!
+npm test
+npm run verify:release
 ```
 
----
-
-## 🙏 Acknowledgments
-
-Built with inspiration from:
-- AI coding community
-- Open source movement
-- Systematic engineering practices
-- Token efficiency research
-- Quality-first development
-
-Special thanks to all contributors and early adopters!
-
----
-
-## 📞 Support & Contact
-
-- **GitHub Issues:** [Report bugs](https://github.com/jeet1511/telo3/issues)
-- **GitHub Discussions:** [Ask questions](https://github.com/jeet1511/telo3/discussions)
-- **Email:** [Contact form](https://github.com/jeet1511)
-- **Sponsor:** [Support development](https://github.com/sponsors/jeet1511)
-
----
-
-<div align="center">
-
-### 🚀 Transform Your AI Coding Today
-
-**Telo3: From Vibe Coding to Systematic Engineering**
-
-Made with ❤️ by [Jeet](https://github.com/jeet1511)
-
-[⭐ Star on GitHub](https://github.com/jeet1511/telo3) | [📖 Read Docs](https://github.com/jeet1511/telo3/blob/main/SKILL.md) | [🤝 Contribute](https://github.com/jeet1511/telo3/blob/main/CONTRIBUTING.md)
-
-</div>
-
----
-
-**Keywords for SEO:** AI coding, AI agent, systematic engineering, project context, AI memory, token optimization, caveman mode, anti-vibe-coding, security framework, privacy compliance, legal awareness, accessibility WCAG, SEO optimization, performance audit, design system, Kiro skill, Cursor extension, Claude integration, ChatGPT plugin, GitHub Copilot, smart delegation, MCP integration, framework-agnostic, production-ready, open-source skill, developer productivity, code quality, AI best practices, Jeet, jeet1511
-
-
----
-
-<details>
-<summary>📊 Project Statistics & Information</summary>
-
-<br/>
-
-![GitHub stars](https://img.shields.io/github/stars/jeet1511/telo3?style=social&label=Star%20this%20AI%20coding%20framework)
-![GitHub forks](https://img.shields.io/github/forks/jeet1511/telo3?style=social&label=Fork%20for%20systematic%20development)
-![GitHub watchers](https://img.shields.io/github/watchers/jeet1511/telo3?style=social)
-![GitHub contributors](https://img.shields.io/github/contributors/jeet1511/telo3)
-![GitHub last commit](https://img.shields.io/github/last-commit/jeet1511/telo3)
-![GitHub repo size](https://img.shields.io/github/repo-size/jeet1511/telo3)
-![Lines of code](https://img.shields.io/tokei/lines/github/jeet1511/telo3)
-![GitHub language count](https://img.shields.io/github/languages/count/jeet1511/telo3)
-
-### Common Search Terms Leading to Telo3
-
-This project helps developers searching for: AI coding assistant frameworks, Claude development tools, GPT-4 project context management, GitHub Copilot productivity enhancement, Cursor AI configuration rules, systematic software engineering approaches, automated project documentation systems, code quality assurance frameworks for AI agents, security compliance in AI-generated code, accessibility standards for software development, performance optimization techniques, SEO best practices, token efficiency in AI communication, empty project setup automation, intelligent onboarding systems, technology stack recommendation engines, smart skill delegation patterns, MCP server integration guides, anti-vibe-coding methodologies, hallucination prevention in AI coding, and production-ready development templates.
-
-### Supported Technologies
-
-**AI Assistants:** Claude (Anthropic), GPT-4 (OpenAI), ChatGPT, GitHub Copilot, Cursor, Kiro IDE, Windsurf, Cody, Tabnine, Amazon CodeWhisperer
-
-**Programming Languages:** JavaScript, TypeScript, Python, Go, Rust, Java, C#, PHP, Ruby, Swift, Kotlin, C++, Scala, Elixir, Haskell
-
-**Frameworks:** React, Vue, Angular, Svelte, Next.js, Nuxt, Remix, Astro, SolidJS, Qwik, Express, Fastify, NestJS, Django, Flask, FastAPI, Rails, Laravel, Spring Boot, ASP.NET
-
-**Development Environments:** VS Code, Visual Studio, IntelliJ IDEA, WebStorm, PyCharm, GoLand, RustRover, Fleet, Zed, Neovim
-
-### Quality Standards Implemented
-
-Telo3 enforces industry-leading quality standards including WCAG 2.1 Level AA accessibility compliance, OWASP security best practices, Core Web Vitals performance optimization, GDPR privacy compliance, COPPA child safety requirements, CAN-SPAM email regulations, DMCA copyright compliance, semantic HTML5 standards, responsive design principles, progressive enhancement methodology, and systematic software engineering practices.
-
-### Use Cases
-
-- **Startup MVPs:** Rapid prototyping with production-quality foundations
-- **Enterprise Applications:** Systematic development with audit trails
-- **Open Source Projects:** Consistent quality across contributors  
-- **Learning Projects:** Best practices built-in from day one
-- **Client Work:** Professional standards, documented decisions
-- **Team Collaboration:** Shared context, reduced onboarding
-- **Solo Development:** Persistent memory across sessions
-- **Migration Projects:** Document existing systems systematically
-- **API Development:** Consistent patterns and security
-- **Full-Stack Applications:** Coordinated frontend/backend development
+The second command packs locally, checks the allowlist, installs the archive offline, and tests the installed CLI, durable skills, hooks, links, idempotency and token baseline. `prepublishOnly` runs both checks. It never publishes or queries credentials.
 
-</details>
+MIT © Jeet ([@jeet1511](https://github.com/jeet1511)).