by lout33
A symbiotic AI agent that remembers everything, challenges you, and extends your cognition.
# Add to your Claude Code skills
git clone https://github.com/lout33/ai-life-assistantA symbiotic AI agent that remembers everything, challenges you, and extends your cognition. Works across Claude Code, opencode, nanobot, and any LLM interface.
Four files that turn any AI into a symbiotic agent with persistent memory, pattern recognition, and autonomous execution. Not a chatbot. An agent that lives in your filesystem and operates alongside you.
curl -fsSL https://raw.githubusercontent.com/lout33/ai-life-assistant/main/install.sh | bash
Or clone and open with your AI coding tool:
git clone https://github.com/lout33/ai-life-assistant
cd ai-life-assistant
| File | Purpose | Stability |
|------|---------|-----------|
| AGENTS.md | Operations, rules, session protocol | Stable |
| SOUL.md | Agent personality, identity | Stable |
| USER.md | Your profile, psychology, patterns | Stable |
| NOW.md | Current state, queue, memory log | Dynamic |
The agent reads all four at session start, updates NOW.md as you work.
| File | Purpose |
|------|---------|
| IDEAS.md | Capture ideas on the fly |
| PROJECTS.md | Detailed project tracking |
| PEOPLE.md | Network and relationships |
| JOURNAL.md | Longer-form reflections |
| LOG_ARCHIVE.md | Archived memory logs |
Just create the file. The agent discovers and uses it.
| Command | What it does |
|---------|--------------|
| /start-day | Morning kickoff. Sets MIT for the day. |
| | Quick accountability check-in. |
| | Evening review. Captures wins, lessons. |
| | Deep reflection. Surfaces patterns. |
No comments yet. Be the first to share your thoughts!
/check-day/end-day/reflectCommands work manually or scheduled via cron. See commands/README.md.
One directory, multiple interfaces:
| Framework | Config |
|-----------|--------|
| Claude Code | ~/.claude/CLAUDE.md (concatenate the 4 files) |
| opencode | ~/.config/opencode/ or project root |
| nanobot | Set workspace in ~/.nanobot/config.json |
Use Claude Code for deep coding, opencode for terminal, nanobot for Telegram/cron. All reading from the same files.
It challenges you. From a real conversation:
AI: "You find something valuable -> People want it -> You feel repulsed by the exchange -> You give it away for free -> You have no money -> Repeat. That's not idealism. That's self-punishment."
It remembers. Persistent memory across sessions. Patterns, quotes, history.
It acts. Writes code, researches, creates files. Not just advice.
In NOW.md, the agent maintains a dated log of patterns, quotes, and insights:
### Jan 10
- Avoided user call. Rescheduled twice. Pattern: building = safe, talking = scary.
### Jan 15
- Had first user call. Quote: 'I've been building what I think they want instead of asking'
### Feb 1
- Pattern confirmed: 3 weeks on feature nobody asked for. This is the 3rd time.
The longer you use it, the better it gets.
Symbiotic > Assistive. Challenge > Validate. Memory compounds. Ship ugly.
| Project | What it does | |---------|--------------| | what-did-i-do | Passive screen tracker with Gemini Vision | | writing-style-skill | Make AI write like you |
Created by @lout33
