# Frequently Asked Questions (FAQ) ❓ Find answers to common questions about **MalikClaw** architecture, installation, hardware requirements, security sandboxing, local LLMs, and mobile automation. --- ## ❓ 1. General & Vision ### Q: How is MalikClaw different from Python agent frameworks like AutoGen, CrewAI, or LangChain? **A:** Most Python agent frameworks consume 200MB–500MB+ of RAM, require heavy virtual environments, and take 10+ seconds to start up. MalikClaw is compiled natively in **Go**. It consumes **<10MB RAM**, boots in **<1s**, ships as a single static binary (~30MB), and runs efficiently on $10 edge hardware (Orange Pi Zero, Raspberry Pi, old Android phones). ### Q: What hardware can run MalikClaw? **A:** MalikClaw runs on almost any computer or embedded board produced in the last 15 years: - **Low-Cost Linux SBCs**: Orange Pi Zero, Raspberry Pi Zero W / 2 W, Sipeed LicheeRV RISC-V. - **Android Smartphones**: Runs natively inside **Termux** (Android 8.0+, non-root). - **Desktops & Cloud VPS**: Linux (`x86_64`, `arm64`), macOS (Apple Silicon / Intel), Windows 10/11, Docker, Kubernetes. --- ## ❓ 2. Installation & Runtime ### Q: Why does terminal say `command not found: malikclaw` after installation? **A:** The installation directory (`$HOME/.local/bin` or `/usr/local/bin`) might not be in your environment `$PATH`. Add it to your shell configuration (`~/.bashrc` or `~/.zshrc`): ```bash export PATH="$HOME/.local/bin:$PATH" echo 'export PATH="$HOME/.local/bin:$PATH"' >> ~/.bashrc source ~/.bashrc ``` ### Q: How do I run MalikClaw as a 24/7 background daemon? **A:** You can launch the gateway server using Docker, systemd, or nohup: ```bash # Via malikclaw gateway command nohup malikclaw gateway > ~/.malikclaw/gateway.log 2>&1 & # Or via Docker docker run -d --name malikclaw --restart unless-stopped -p 18790:18790 -v ~/.malikclaw:/root/.malikclaw ghcr.io/abdullahmalik17/malikclaw:latest ``` --- ## ❓ 3. LLM Providers & Local Models ### Q: How do I run 100% offline private LLMs with Ollama? **A:** Install Ollama (`ollama serve`) and pull your model (e.g. `ollama pull llama3.3`). In `~/.malikclaw/config.json`, add Ollama to `model_list`: ```json { "model_list": [ { "model_name": "llama3.3", "model": "ollama/llama3.3", "api_base": "http://localhost:11434/v1" } ] } ``` ### Q: How do automated fallback chains work? **A:** If your primary LLM API key runs out of quota, hits rate limits (HTTP 429), or suffers downtime, MalikClaw automatically retries down the ordered list of models defined in `model_list`. --- ## ❓ 4. Security & Sandboxing ### Q: Why did MalikClaw refuse to run my command (`Command blocked by security filter`)? **A:** MalikClaw enforces strict regex filtering on shell execution to prevent privilege escalation (`sudo`, `su`), catastrophic file deletion (`rm -rf /`), disk formatting (`mkfs`), or dangerous remote script pipes (`curl | sh`). ### Q: Can I allow file access outside `~/.malikclaw/workspace`? **A:** Yes. You can configure path allowlists in `~/.malikclaw/config.json`: ```json { "tools": { "allow_read_paths": ["/var/log/syslog", "/etc/hosts"], "allow_write_paths": ["/home/user/backups/"] } } ``` --- ## ❓ 5. Channels & Mobile Control (ADB) ### Q: How do I connect my Android phone for ADB control? **A:** 1. Enable **Developer Options** and **USB Debugging** on your phone. 2. Connect your phone to your computer via USB or Wi-Fi ADB. 3. Verify connection by running `adb devices`. 4. Issue natural language commands to MalikClaw: `malikclaw agent -m "Take a screenshot of my phone"`. ### Q: How do I prevent strangers from talking to my agent on Telegram or Discord? **A:** Set the `allow_from` whitelist in your channel configuration inside `config.json` with your specific user IDs: ```json { "channels": { "telegram": { "enabled": true, "bot_token": "YOUR_BOT_TOKEN", "allow_from": ["@your_telegram_username", "123456789"] } } } ``` --- ## ❓ 6. Troubleshooting & Diagnostics ### Q: Where are log files located? **A:** Standard logs are output to stdout/stderr. When running the gateway daemon, logs are saved in `~/.malikclaw/logs/` or accessible via: ```bash malikclaw logs --follow ``` ### Q: How do I enable debug logging? **A:** Pass the `--verbose` flag on CLI commands: ```bash malikclaw agent -m "Debug system state" --verbose ```