Introduction and Glossary — What AI agents are, what ClawNanny is, an honest assessment of AI agents' probable impact on the labor economy, and definitions of terms used throughout this guide Choose your agent backend — OpenClaw vs. Hermes Agent vs. Claude Code (OAUTH subscription) — what differs, how to switch, what happens to your data, which ClawNanny features are backend-specific UI:   Interact with your agent to assign it work — Sending messages, reading replies, Talk mode, session switching, image pasting, the context indicator, stopping the agent UI:   You're in charge! — Visibility, stopping the agent, scheduled task safeguards, skill consent, budget stops, your responsibilities — a direct answer to "What keeps it from going off the rails?" UI:   Sessions, focal points, context & memory — Managing your agent's current state of mind, a session as a conversation thread with a limited scope, starting and renaming a session, understanding the "context window" (knowledge basket), viewing a transcript, saving ("persisting") memories that sessions can share (continual learning) UI:   Knowledge capture — How memories are stored and retrieved; how you can view, search, and edit memory files UI:   Configuration — Adjusting your agent's behavior, form view vs. JSON editor, saving and applying changes UI:   Scheduled tasks — Setting up recurring tasks ("cron jobs"), reviewing run history, pausing and deleting tasks, practical guidelines for unattended automation UI:   Skills — What skills are, finding and installing them, the security scan, enabling and disabling skills, providing API keys for skills that need them UI:   Context state — How to view the state of the context window ("knowledge basket") for the current session UI:   Messaging channels — Connecting Telegram, WhatsApp, and other messaging apps so you can reach your agent without opening the dashboard UI:   Manage token costs — How AI billing works, reading cost summaries, setting budget limits UI:   Tokens & API keys — Managing API keys and authorization tokens UI:   Security — Ensuring the security of your data assets and protections against intrusion UI:   Health checks — A summary page showing the status of all of ClawNanny's system health checks Bot manager:   Local — Starting, stopping, and upgrading OpenClaw; connecting to a remote instance; workspace file management Bot manager:   Monitor security — How ClawNanny continuously monitors security Bot manager:   My devices — Which devices can connect with your agent Bot manager:   Test services — Prove which services are available Bot manager:   Microsoft Outlook — Configure how your agent interacts with Microsoft Outlook Bot manager:   Remote access — Managing remote access to services Bot manager:   System diagnostics — System diagnostics, the agent's console log, and connection Log — what and how to check when something isn't working Language models — Configure preferences — Manage which language models you interact with and your priorities for each one Workspace configuration files — SOUL.md, AGENTS.md, USER.md, etc. System Markdown documents — System Markdown documents My Markdown documents — My Markdown documents (as opposed to the shared docs you can't edit) Agent persona builder — How you give your agent one or more personalities to choose from, using sliders Audit configuration and security — Monitoring all types of vulnerability to intrusion by unauthorized actors, naming specific attack surfaces, describing (or directly implementing) solutions Audit for enterprise-level security — Configuration audit and enterprise-level security audit — what they check, how to read results, exporting reports Common Questions — FAQ — Plain-English answers to the questions new clients ask most often Telegram: one bot per device — Why Telegram messages stop arriving (or arrive twice) when two devices share one bot token, and how to fix it — the *409 Conflict* coaching page

File-system authorizations — Authorize agent operations, folder by folder ClawNanny Governance Policy (CGP)

Language-model list Add a language model to the list Language-models configuration — Configure your model order preferences Language-model eval runs — Test models against each other

Time zones — View corresponding times in two time zones Markdown to HTML — Convert Markdown text to HTML HTML to Markdown — Convert HTML text to Markdown JSON parser — View a JSON string as a hierarchical outline

Create a new account — Create a new client account for access to the ClawNanny mother ship Client account management — Manage my client account on the ClawNanny mother ship Reset my password — Reset my account password on the ClawNanny mother ship. Cancel my subscription — Stop recurring charges for ClawNanny.

Monitor your computer's vitals — CPU, memory Database connections monitor — View the database connection pool Detect MCP servers — Detect installed MCP servers Internet speed test

OpenRouter API expenses incurred OpenRouter API call logs Anthropic subscription tokens consumed Anthropic API expenses incurred OpenAI Codex subscription tokens consumed OpenAI API expenses incurred OpenAI credit balance

Sessions search — Search topics across sessions Sessions — OpenClaw — Search OpenClaw sessions Sessions — Hermes — Search Hermes sessions Sessions — ClaudeCode — Search Claude Code sessions Sessions — Codex — Search Codex sessions

OpenClaw configuration — Edit the OpenClaw configuration OpenClaw logs — View logs about OpenClaw's behavior Token operating costs — OpenClaw — Token costs for OpenClaw

Token operating costs — Hermes (MISSING) — Token costs for Hermes