# Autonomous Operator > Autonomous Operator Skill - Skill: `koinod/autonomous-operator` (Agent Skill, multi-file: 4 files) - Install (CLI): `npx skillmds@latest add koinod/autonomous-operator` - Raw SKILL.md: https://api.skillmd.com/api/skills/koinod/autonomous-operator/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Coding & Dev Tools - Author: koinod (https://skillmd.com/u/koinod) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/koinod/autonomous-operator --- # Autonomous Operator Skill Provides autonomous business operation capabilities: multi-backend email, local LLM integration, self-monitoring, credential management, and revenue tracking. Enables the agent to run without manual gatekeepers. ## Capabilities - **Multi-backend email** — automatically detects and uses available email methods (gog OAuth, SMTP app password, third-party APIs) - **Local LLM** — integrates with Ollama for unlimited tokens, no API costs - **Self-monitoring** — health checks, revenue pipeline tracker, alerting - **Credential vault** — secure storage and rotation of service credentials - **Autonomous loops** — periodic execution of high-value tasks (follow-ups, research, monitoring) ## Triggers - "Set up autonomous email" - "Enable local LLM" - "Start self-monitoring" - "Build sovereign business machine" - "Remove gatekeepers" ## What It Does ### 1. Email Independence Detects available email sending methods in order of preference: 1. gog CLI with OAuth (if configured) 2. SMTP with app password (if credentials available) 3. Third-party email API (SendGrid, Mailgun, etc.) 4. Fallback: draft emails for manual send Provides unified interface: `send-email --to --subject --body ` Automatically retries with fallback if primary fails. ### 2. Local LLM Integration Installs and manages Ollama on the local machine: - Detects if Ollama is installed, installs if not (via brew or direct binary) - Pulls Llama 3.2 (or other models) based on available RAM - Configures OpenClaw to use local LLM for agent instructions - Provides unlimited token generation for agent workflows ### 3. Self-Monitoring Dashboard Generates daily status reports: - Agent health (running, errors, restarts) - Revenue pipeline (leads, emails sent, replies, calls booked, deals closed) - Credential status (expiring soon, needs rotation) - System resources (RAM, CPU, disk) - Upcoming tasks and follow-ups Dashboard can be viewed via: `openclaw status --full` or a simple web UI. ### 4. Autonomous Loops Runs scheduled tasks without cron (internal scheduler): - Every 2 hours: check inbox, send scheduled follow-ups, update lead status - Daily: generate revenue report, update next-moves.md, log insights - Weekly: competitor watch, pricing intel, template refresh - On trigger: when new lead added, automatically enrich and add to sequence ### 5. Credential Vault Securely stores: - Email passwords/app passwords - API keys (OpenRouter, etc.) - Client credentials - Rotation reminders Uses macOS Keychain or encrypted file storage. ## Setup ```bash # Initialize autonomous operator openclaw sessions_spawn --runtime subagent --task "Initialize autonomous operator: install Ollama, configure email fallbacks, set up monitoring" # Or use the CLI directly (once installed) autonomous-operator init autonomous-operator email-test autonomous-operator dashboard ``` ## Configuration Environment variables: - `AUTONOMOUS_EMAIL_BACKENDS` — comma-separated list of backends to try (default: gog,smtp,api,draft) - `AUTONOMOUS_LLM_MODEL` — local model to use (default: llama3.2) - `AUTONOMOUS_REPORT_TIME` — daily report time (default: 09:00) - `AUTONOMOUS_FOLLOWUP_INTERVAL` — hours between follow-up cycles (default: 2) ## Benefits - No dependency on any single credential provider - Works offline (local LLM) for core agent instructions - Automatic failover ensures continuity - Self-monitoring means you know status at a glance - Reduces manual repetitive work to zero ## Notes This skill is designed for sovereignty. It assumes you want to own your infrastructure and not rely on third-party APIs that can go down or change terms. Local LLM gives you unlimited tokens and privacy. Preferred hardware: Mac mini or Linux server with at least 16GB RAM for comfortable Llama 3.2 operation.