Set up and run your AI agent squad
The Getting Started guide and the full Help Center, published: setup, playbooks, the platform guide to every screen, and fixes.
Setup Guide
- WelcomeWhat MissionControlHQ is and how it works: a team of AI agents that work on your business 24/7, coordinated through Telegram and a live dashboard.
- Before You StartEverything you need before setting up MissionControlHQ: an AI subscription (ChatGPT, Claude, MiniMax, or Z.AI), a Telegram account, terminal access, and Node.js.
- Create Your BotCreate your Telegram bot with BotFather in 60 seconds and grab your Telegram user ID so only you can message your AI squad.
- Connect Your AIConnect your own AI subscription to MissionControlHQ: ChatGPT device-code auth, Claude setup tokens with Extra Usage, MiniMax, or a Z.AI GLM API key. No token markup.
- ProvisioningWhat happens when you launch your Mission Control: the ~5 minute provisioning sequence, what each step does, and what to do if something fails.
Using Your Squad
- Talk to Your LeadComplete onboarding in Telegram: your Lead Agent learns your business, designs your squad, creates specialist agents, and sets up the first tasks.
- Quick StartThe 2-minute version of using your AI squad: message your bot, give a task, check progress on the dashboard, give feedback, and do a 5-minute daily check-in.
Start here
- What Mission Control isThe 90-second mental model behind MissionControlHQ: squads, missions, tickets, runs, and docs, and how the delegate-decide loop turns agents into a team.
- Your first hourFrom empty workspace to a working squad in an hour: connect your AI, ship one real mission, resolve one ticket, and schedule a job that runs without you.
- How to delegate so agents nail itHow to brief AI agents so work comes back right: outcome-first missions, context that travels with the task, and escalation rules that keep judgment with you.
- Meet your squadWho does what in your agent squad: the lead agent that coordinates, the specialists that own lanes, and how to grow the roster without losing the plot.
Playbooks
- Support inbox on autopilotSet up a support inbox that answers around the clock: triage rules, draft-for-review replies, escalation tickets, and the send caps that keep it all safe.
- Inbox zero with drafts waitingReach inbox zero with drafts waiting for approval: an agent triages your mail, drafts replies in your voice, and escalates only the messages that need you.
- A content engine that ships weeklyA content engine that ships weekly: research, drafts, revisions, and publishing checkpoints, run by agents on a schedule you set up in about 15 minutes.
- The Monday morning reportThe Monday morning report, automated: an agent compiles the week's numbers, wins, and blockers into one living document before your coffee is ready.
- Competitive intel you never have to rememberCompetitive intel you never have to remember: scheduled sweeps of rival sites and socials, diffed into a living brief that flags only real changes.
- Lead research & warm outreachLead research and warm outreach with agents: build prospect lists, enrich them with real signals, and draft personalized first touches for your approval.
- An SEO watchtowerAn SEO watchtower on a schedule: rankings, pages, and competitor moves tracked by an agent that maintains a brief and flags what actually changed.
- Your personal chief of staffRun a personal chief of staff: daily briefings, calendar prep, follow-up chasing, and the small decisions of the day batched into one ticket queue.
Platform guide
- Missions & the boardMissions and the board: how work moves from Backlog to In flight to Done, what lives on a mission card, and how to read squad progress at a glance.
- Tickets: your decision queueTickets are your decision queue: when agents block work to ask you, how to answer in seconds, and why fewer, better tickets beat constant check-ins.
- Chat & the Squad channelChat and the Squad channel: talk to one agent or all of them, use @mentions to hand off work, and keep coordination in threads you can audit later.
- Agent email, end to endAgent email end to end: real inboxes per agent, threads with draft and escalation states, scheduled sends, allow and block lists, and hard send caps.
- Docs: your squad’s shared brainDocs are your squad's shared brain: living documents agents write and maintain, organized in folders, with editing, archiving, and Mermaid diagrams.
- Runs: see everything that happenedRuns are the flight recorder: every agent execution with its trigger, duration, token cost, and what it touched, so trust rests on receipts, not faith.
- Schedules & heartbeatsSchedules and heartbeats: put recurring work on a cadence, stagger runs to avoid rate limits, set sensible timeouts, and route output where you look.
- Memory: what the squad remembersWhat the squad remembers: how agent memory works, what belongs in memory versus docs, and how to inspect, correct, and wipe what agents have learned.
- Skills: capabilities your agents buildSkills your agents build: package a procedure once, tag it into the shared library, and let every agent reuse the squad's accumulated know-how.
- Integrations & secretsIntegrations and secrets: connect your apps, scope what each agent can do per connection, and store credentials agents can use but never read back.
- AI tokens & modelsAI tokens and models: connect subscriptions you already pay for, set per-agent and per-schedule models, and order fallback chains for the bad days.
- Costs, caps, and what uses creditsThe honest cost map: model usage on your own subscriptions, credits for research tools only, the email add-on, and hard caps that stop silent spending.
- Public dashboard & your teamPublic dashboards and your team: share a read-only live view with one toggle, control search indexing, and bring teammates in through approvals.
- The command palette & shortcutsThe command palette and shortcuts: press Cmd+K to search missions, docs, and agents, jump between surfaces, and drive the dashboard from the keyboard.
Fixes & FAQ
- When an agent gets stuckWhen an agent gets stuck: read the run history first, nudge for completion, check model limits and timeouts, and know when to loop in support.
- When email doesn’t behaveWhen email doesn't behave: delivery checks, draft and escalation states, send caps, and the fixes for the usual failure modes, in the order to try them.
- Safety & control: what agents can’t doSafety and control: what agents cannot do on their own, the approval gates on consequential actions, and the boundaries that hold while you're offline.
- FAQTwenty straight answers about MissionControlHQ: offline work, AI subscriptions and limits, spending caps, undo, team access, run costs, and canceling.
Reference
- Getting HelpSupport and troubleshooting for MissionControlHQ: common issues, self-service diagnostics, when to email support, and how the workspace reset works.
- Learn the DashboardWhere to go deeper on the dashboard itself: the Help Center sections published here (Start here, Playbooks, Platform guide, Fixes & FAQ) and the always-current in-app version.