Step-by-step guides to get the most out of your AI agent. No coding required.
Set your expectations right. Everything your agent can do out of the box, what needs setup, and what's not possible yet.
The single most impactful thing you can do. Give your agent a name, a voice, and a personality with ready-to-use templates.
A fast diagnosis flow for when personality edits seem ignored—verify file writes, test the right surface, and force a clear style change.
What models are, what they cost, how to get API keys, and which model is right for you.
Step-by-step setup for OpenRouter, plus beginner-friendly advice on which free models to try first.
Talk to your AI agent from any Discord server. Full walkthrough plus troubleshooting for the most common issues.
Set up your agent on Telegram in under 5 minutes. The easiest channel to get running.
Configure key + voice ID correctly, avoid free-tier gotchas, and run a known-good voice smoke test in minutes.
Why your agent "forgets" things, how memory actually works under the hood, and how to make important things stick.
A simple privacy audit flow to verify what your agent can access, what it remembers, and how to reduce oversharing risk.
Where file attachments work, where they don't, and workarounds for getting images and documents to your agent.
Stop screenshot hallucination loops with a fast visibility test, recovery flow, and support-ready escalation packet.
A plain-English guide to session length, bloated chats, lost work, and the 4 habits that keep your progress safe.
A fast recovery pattern for long-chat degradation: preserve what matters, use a compact handoff, and avoid context bloat loops.
A fast recovery flow for post-reset silence, including clean-surface ping tests, permission checks, and session-mismatch fixes.
Complete guide to the OpenClaw Control UI. What every page, button, and field does — so you never feel lost.
Why this is usually expected, how to verify your setup in 2 minutes, and when it’s a real issue worth escalating.
A plain-English guide to approval prompts: what they mean, when to approve, when to deny, and how to avoid getting stuck.
A fast fix for config save failures caused by redacted placeholder values being treated as real secrets.
A practical guide to getting better outputs from your agent using a simple 4-part prompt formula and real before/after examples.
A fast diagnostic checklist for laggy replies, long-running tasks, approval pauses, and getting back to reliable speed.
Use a quick liveness + execution isolation flow to confirm runtime issues and recover safely without guessing.
Fix execution-mismatch trust breaks with a simple “done + evidence + verification” workflow and escalation packet.
A practical recovery checklist for common 502 failures, plus a clean support packet you can paste into #help.
A quick isolation flow for blank dashboard pages caused by browser session, extension, VPN/proxy, or network-path issues.
Stop over-chatty behavior in Discord with clear response gates, silence rules, and a copy/paste group-chat policy block.
A focused recovery checklist for when “/” commands don't appear, including role permissions, channel overrides, and known-good tests.
A practical fix for guild/channel ID mixups and hidden channel permission overrides that make bots look "broken."
Use a clean-room channel test to isolate send-permission, thread/forum, and integration restriction issues in minutes.
Fix the high-friction typing-then-silent loop with a fast channel-policy isolation flow and known-good recovery steps.
A plain-English fix for the common confusion where DM works but server channels feel like a different agent.
Fix the “it forgot everything” confusion when switching between dashboard chat, Discord DMs, and server channels.
A fast path-first workflow to find missing outputs, verify actual write location, and stop "saved but can’t find it" loops.
Fix Discord DM privacy settings so Whambulance can open your support flow and your ticket doesn’t get stuck.
Fix repetitive heartbeat behavior with clean checklist design, non-repetition rules, and lightweight state tracking.
A practical recovery guide for heartbeat file path mistakes, stale instructions, repeated loops, and noisy check-ins.
Fix timezone drift that causes wrong-day replies, off-schedule reminders, and confusing time-based behavior.
A practical 5-minute diagnosis flow for missing/late reminders, timezone mismatches, and wrong delivery context.
A plain-English guide to the most common confusion: skills are behavior packs, not native software installs.
Stop unnecessary setup loops with a no-install recovery prompt, assumption check, and safer decision flow.
Fix the common “needs CLI” confusion with a connector/auth check, tiny write test, and clean escalation packet.
Understand read-only web access vs browser automation, diagnose capability limits fast, and use a clean human+agent handoff.
Diagnose private/login-gated URLs, anti-bot blocks, and safe workarounds when your browser can see pages the agent cannot.
A practical fix for “Dropbox MCP required” errors, with connector/auth checks, smoke test prompt, and escalation packet.
Fix connector "installed but unavailable" loops with a fast auth/session reset flow, smoke test prompt, and clean escalation packet.
Pre-built SOUL.md files for every use case — personal assistant, copywriter, fitness coach, real estate agent, and 48 more. Copy, paste, go. Free forever.
15+ ready-to-use skills: human writing, market research, daily briefings, email management, SEO, journaling, and more. Install in one command.
Showcase spotlights, community tips, the fox phenomenon, common issues, and shoutouts to the people keeping this community alive.