The World's Most Advanced AI Assistant — Self-Improving, Multi-Agent, Knowledge-Powered
Built by Friday 🤖 — Production-tested, self-improving AI assistant template
Friday is a production-ready AI assistant built on OpenClaw with:
- ✅ Self-Improvement Loop — Learns from every interaction
- ✅ Multi-Agent System — 5+ specialized agents working together
- ✅ LLM Wiki — Karpathy-inspired knowledge base
- ✅ Meeting Transcription — Whisper-powered auto-notes
- ✅ Wellness Tracking — Mood, sleep, productivity correlation
- ✅ No-Code Agent Creator — Create custom agents via wizard
- ✅ 20+ Examples — Copy-paste usage patterns
- ✅ Crew-Inspired UX — Best features from My Brain Is Full Crew
Result: An AI assistant that gets smarter every day, documents itself, and actually helps you get things done.
# Install OpenClaw
curl -fsSL https://openclaw.ai/install.sh | bash
# Install Node.js (v22+)
nvm install 22
# Install Python (3.11+)
brew install python@3.11# Clone Friday's workspace
git clone https://github.com/YOUR_USERNAME/friday-ai-blueprint.git
cd friday-ai-blueprint
# Copy to your OpenClaw workspace
cp -r . ~/.openclaw/workspace/
# Bootstrap Friday
cd ~/.openclaw/workspace
./bootstrap.sh# Start OpenClaw
openclaw start
# Friday will automatically:
# 1. Read SOUL.md, USER.md, IDENTITY.md
# 2. Load memory files
# 3. Check tasks/todo.md
# 4. Run heartbeat checksfriday-ai-blueprint/
├── 📄 README.md # This file
├── 📄 SOUL.md # Friday's personality & rules
├── 📄 USER.md # About your human (Roberto)
├── 📄 IDENTITY.md # Friday's identity & capabilities
├── 📄 AGENTS.md # Workspace conventions
├── 📄 HEARTBEAT.md # Background task schedule
├── 📄 TOOLS.md # API keys & tool configs
│
├── 📂 lab/ # Core automation scripts
│ ├── trace_logger.py # Log agent executions
│ ├── trace_analyzer.py # Find patterns in traces
│ ├── meta-improver.py # Self-improvement engine
│ ├── monitor_deployments.py # Monitor agent performance
│ ├── verify_improvement.py # Verify improvements work
│ ├── scribe.py # Meeting transcription
│ ├── wellness_tracker.py # Mood + sleep tracking
│ ├── agent_creator.py # No-code agent wizard
│ ├── daily_task_prep.py # Daily task preparation
│ ├── email_triage.py # Email prioritization
│ ├── calendar_check.py # Calendar management
│ └── morning_brief.py # Morning briefings
│
├── 📂 skills/ # Friday's capabilities
│ ├── agentshield/ # Security scanning
│ ├── pageindex/ # Document analysis
│ ├── scribe/ # Meeting transcription
│ ├── wellness-tracker/ # Wellness tracking
│ ├── agent-creator/ # Agent creation
│ ├── executive-assistant/ # Executive tasks
│ └── daily-task-prep/ # Task preparation
│
├── 📂 friday-wiki/ # LLM knowledge base
│ ├── wiki/ # Compiled knowledge
│ │ ├── agents/ # Agent profiles (8+)
│ │ ├── skills/ # Skill docs (15+)
│ │ └── concepts/ # Auto-discovered concepts
│ ├── examples/ # Usage examples (20+)
│ ├── tools/ # Wiki tools
│ │ ├── wiki-compile.py # Compile raw → wiki
│ │ ├── wiki-search.py # BM25 search
│ │ ├── wiki-qa.py # Q&A system
│ │ ├── wiki-lint.py # Health checks
│ │ └── wiki-render.py # Marp/chart rendering
│ └── templates/ # Agent templates
│
├── 📂 memory/ # Persistent memory
│ ├── YYYY-MM-DD.md # Daily logs
│ ├── wellness/ # Wellness tracking
│ └── entities/ # Entity memories
│
├── 📂 tasks/ # Task management
│ ├── todo.md # Active tasks
│ ├── lessons.md # Lessons learned
│ └── WORKFLOW.md # Workflow reference
│
└── 📂 configs/ # Configuration files
├── governance.json # Meta-improver policy
└── agents/ # Agent definitions
Friday learns from every interaction:
Agent executes → Logs trace → Analyzes patterns →
Proposes improvement → Debates quality →
Gets approval → Deploys → Monitors → Verifies → Repeat
Files:
lab/trace_logger.py— Log executionslab/meta-improver.py— Generate improvementslab/monitor_deployments.py— Monitor 24hlab/verify_improvement.py— Verify results
Usage:
# Check improvement status
python3 lab/meta-improver.py --status
# View trace history
python3 lab/trace_analyzer.py --summaryFriday coordinates 5+ specialized agents:
| Agent | Purpose | Status |
|---|---|---|
| Friday | Personal assistant | 🟢 Active |
| OpenSpace | Research & exploration | 🟢 Active |
| Nightshift | Overnight automation | 🟢 Active |
| Shannon | Security pentesting | 🟢 Active |
| clawchief | Business workflows | 🟢 Active |
| Scribe | Meeting transcription | 🟢 Active |
| Wellness Coach | Mood tracking | 🟢 Active |
Skills Hub: Agents share skills via publisher/subscriber pattern.
Self-documenting knowledge base:
Raw Data (traces, proposals) →
LLM Compiler → Wiki Articles →
Q&A System → Auto-File Outputs →
Enhance Wiki → Repeat
Tools:
wiki-compile.py— Compile raw data → wikiwiki-search.py— BM25 search enginewiki-qa.py— Answer complex questionswiki-lint.py— Health checks + suggestionswiki-render.py— Marp slides + charts
Usage:
# Ask question
python3 friday-wiki/tools/wiki-qa.py "What agents exist?" --output
# Search wiki
python3 friday-wiki/tools/wiki-search.py "meta-improver"
# View in Obsidian
open ~/.openclaw/workspace/friday-wikiWhisper-powered meeting automation:
# Transcribe meeting
python3 lab/scribe.py --transcribe meeting.wav --model base
# Extract action items
python3 lab/scribe.py --extract-actions transcript.md
# File in Obsidian
python3 lab/scribe.py --file-in-obsidian transcript.md
# Draft follow-up email
python3 lab/scribe.py --draft-followup transcript.mdTime Saved: ~40 minutes per meeting
Mood, sleep, and productivity correlation:
# Morning check-in
python3 lab/wellness_tracker.py --checkin morning
# Evening check-in
python3 lab/wellness_tracker.py --checkin evening
# Weekly summary
python3 lab/wellness_tracker.py --summary weekly
# Correlation analysis
python3 lab/wellness_tracker.py --analyze --days 30Insights:
- Sleep quality → Energy correlation
- Mood → Productivity patterns
- Stress → Performance impact
Create custom agents via wizard:
# Interactive wizard
python3 lab/agent_creator.py --wizard
# From template
python3 lab/agent_creator.py --template research
# Available templates:
# - research (deep research specialist)
# - wellness (mood tracking)
# - finance (budget management)
# - learning (study planning)Creates:
- Wiki article with YAML frontmatter
- Skill documentation
- Script template
- Configuration JSON
Time: ~30 seconds per agent!
20+ copy-paste usage examples:
| Category | Examples |
|---|---|
| Email Triage | 5 examples |
| Calendar Management | 5 examples |
| Meeting Notes | 5 examples |
| Custom Agents | 5 examples |
Browse:
open ~/.openclaw/workspace/friday-wiki/examples/Best features from My Brain Is Full Crew:
- ✅ YAML agent metadata (standardized definitions)
- ✅ Examples library (learn by doing)
- ✅ Meeting transcription (auto-notes)
- ✅ Wellness tracking (holistic approach)
- ✅ No-code agent creator (empowerment)
Plus Friday's Superpowers:
- 🔥 Self-improvement loop (Crew doesn't have)
- 🔥 Trace system (Crew doesn't have)
- 🔥 Debate + governance (Crew doesn't have)
- 🔥 Multi-agent learning (Crew doesn't have)
- 🔥 Sherlock research (Crew doesn't have)
| Metric | Value |
|---|---|
| Time to Deploy | ~30 minutes |
| Code Created | 128 KB |
| Files | 40+ |
| Agents | 8 |
| Skills | 15+ |
| Examples | 20+ |
| Wiki Articles | 14+ |
| Time Saved | ~20 hours/month |
# Morning (8 AM)
python3 lab/daily_task_prep.py # Prepare tasks
python3 lab/calendar_check.py --hours 24 # Check schedule
python3 lab/wellness_tracker.py --checkin morning # Mood check
# During Day
python3 lab/scribe.py --transcribe meeting.wav # Meeting notes
python3 friday-wiki/tools/wiki-qa.py "Question?" # Ask wiki
# Evening (9 PM)
python3 lab/wellness_tracker.py --checkin evening # Reflection
python3 lab/email_triage.py --max 20 # Inbox zero# Sunday Morning
python3 lab/trace_analyzer.py --summary # Review traces
python3 lab/wellness_tracker.py --summary weekly # Wellness review
python3 friday-wiki/tools/wiki-lint.py --suggest # Wiki health# Run wizard
python3 lab/agent_creator.py --wizard
# Example session:
# 1. Name: research-assistant
# 2. Triggers: research, study, find
# 3. Skills: web-search, summarization
# 4. Languages: en, es
# 5. Priority: 2
# Result:
# ✅ friday-wiki/wiki/agents/research-assistant.md
# ✅ skills/research-assistant/SKILL.md
# ✅ lab/research-assistant.pyEdit TOOLS.md with your keys:
## NVIDIA NIM
- **Key:** `nvapi-xxx`
- **Base URL:** `https://integrate.api.nvidia.com/v1`
## Alibaba DashScope
- **Key:** `sk-xxx`
- **Base URL:** `https://dashscope.aliyuncs.com/compatible-mode/v1`
## Exa AI (Web Search)
- **Key:** `xxx`
- **Usage:** Sherlock deep research# Edit crontab
crontab -e
# Add Friday's automation:
# Morning task prep (6 AM daily)
0 6 * * * cd ~/.openclaw/workspace && python3 lab/daily_task_prep.py
# Email triage (every 4 hours)
0 */4 * * * cd ~/.openclaw/workspace && python3 lab/email_triage.py
# Calendar check (every 4 hours)
0 */4 * * * cd ~/.openclaw/workspace && python3 lab/calendar_check.py
# Morning brief (8:30 AM daily)
30 8 * * * cd ~/.openclaw/workspace && python3 lab/morning_brief.py
# Wellness check-ins
0 8 * * * python3 lab/wellness_tracker.py --checkin morning
0 21 * * * python3 lab/wellness_tracker.py --checkin evening
# Meta-improver monitoring
0 * * * * python3 lab/monitor_deployments.py
0 6 * * * python3 lab/verify_improvement.py
30 4 * * 0 python3 lab/trace_analyzer.py
0 5 * * 0 python3 lab/meta-improver.py- Self-Improvement — Read
lab/meta-improver.py - Multi-Agent — Read
lab/MULTI_AGENT_INTEGRATION.md - Wiki System — Read
friday-wiki/README.md - Skills Hub — Read
skills/README.md
- Discord: https://discord.com/invite/clawd
- GitHub: https://github.com/openclaw/openclaw
- ClawHub: https://clawhub.ai (find skills)
MIT License — See LICENSE file.
# 1. Clone this repo
git clone https://github.com/YOUR_USERNAME/friday-template.git
# 2. Copy to OpenClaw workspace
cp -r friday-template ~/.openclaw/workspace/
# 3. Customize for your needs
cd ~/.openclaw/workspace
# Edit USER.md, SOUL.md, IDENTITY.md
# 4. Bootstrap
./bootstrap.sh
# 5. Start using!
openclaw startBuilt with ❤️ by Azazelpy& Friday
friday-ai-blueprint v1.0.0 — The World's Best AI Assistant 🚀
Friday 🤖
Personal AI Assistant, Chief of Staff, Technical Partner
Created: 2026-04-05
Version: 1.0.0
License: MIT
