A Lead Bot in a Grok Bot group is the coordination seat — not an “AI employee” that emails under your name. It breaks a scoped ask into specialist jobs, collects in-box artifacts, and hands the human one review pack. The Lead stops before send, publish, or pay; specialists stop the same way. The Lead owns prioritization, scoped handoffs, and the review queue. Specialists own one artifact type per job and do not widen scope alone. Done is a reviewable artifact pack the human can approve in one pass — not an open-ended chat that might fire a tool later. Keep the roster short so jobs do not collide.
Who this is for
Solo founders and small ops desks already in Cursor + Grok Bot who want a named Lead without “the bot sent something I can’t undo.” If you need the full roster and stop language, start with Grok Bot role templates that stop before send/publish/pay. If you need the handoff loop in depth, use Lead Bot + specialist Bot coordination. This FAQ owns the definition: what Lead *is*, what it owns, and what it never does alone.
Lead vs specialist (jobs, not personas)
| Seat | Owns | Produces | Never does alone |
|---|---|---|---|
| Lead Bot | Prioritization, scoped handoffs, review pack | One consolidated queue for the human | Auto-send / publish / pay |
| Specialist | One job, one artifact type | `draft.md` / checklist / brief in-box | Widen scope; leave the box |
Prefer a short roster (Lead + a few specialists) over one mega-Bot that claims Chief of Staff + writer + publisher. Coordination beats a persona dump. Free “AI employee” template kits can still give you day-one labels — an agent-ops desk adds the gates. Contrast: agent-ops desk vs AI-employee template kits.
Coordination loop (approval-gated)
- Human (or a routine that only *starts* draft work) gives the Lead a scoped ask.
- Lead breaks the ask into specialist jobs: one job, one artifact type, named stop list.
- Specialists produce in-box artifacts — not outbound messages.
- Lead assembles a single review pack.
- Human Approves, edits, or rejects. Only then does send / publish / pay happen — and that press is human (or an explicitly granted send permission), never silent auto-fire.
Specialists do not widen the brief alone. Lead reassigns; specialists stay in the seat they were given. Secrets stay in env / secret stores — never in Bot descriptions, skills, or chat.
What the Lead never owns alone
Put these in the Lead’s Bot description in language the model can’t “creatively reinterpret”:
- Send — email, Slack, SMS, chat: draft only; human Approves before send.
- Publish — blog, social, changelog, store listing: draft artifact; human hits publish.
- Pay — Stripe, invoices, refunds, buys: propose; human executes.
- Delete / share outside the box — stop and ask.
“Done” for a Lead-run job is a reviewable artifact pack the human can approve in one pass — not an open-ended chat that might fire a tool later. Routines may schedule work; they do not grant irreversible rights. For repo agents that share the same desk, mirror stop-before-send/publish/pay in root `AGENTS.md`.
Lead failure card (pin this)
| Symptom | Likely cause | Fix |
|---|---|---|
| Lead sent or posted without asking | Stops missing from Bot description | Put never-send/publish/pay in the description; skills cannot override |
| Specialist “helps” by emailing | Lead delegated an outcome, not an artifact | Job = file/PR/URL draft; send gate stays with human |
| Mega-Bot does everything poorly | One description claims CoS + writer + publisher | Split into Lead + named specialists |
| Specialist widens scope mid-run | No “Lead reassigns” rule | Specialist stays in seat; Lead re-scopes |
| Chat says “done” with nothing to review | Done = conversation | Redefine done as a reviewable artifact pack |
| Secrets appear in Lead/specialist text | Credentials pasted into description/skill | Move to env / secret store; scrub chat surfaces |
Debug order: read Lead stop list → confirm specialist jobs are artifact-shaped → confirm skills have no send/publish/pay side effects → confirm routines only trigger drafts → then add seats.
What Cheat Sheet + Agent Starter are for
The Cheat Sheet is the skimmable roster and approval checklist for Lead + specialist desks — name it in your runbook. Soft path today: start on /free (Harness Desk Starter — seat, done-artifact skill, gate card). When you want harness depth (role pack + agent-ops files, not AI-SDR SaaS), use the Agent Starter Kit ($47).
Stay in agent-ops — never Domo-as-kit-ROI, never “replaces your team,” never free Operator Kit fluff that still auto-sends under your name. Seat templates: Grok Bot role templates…. Desk vs template-kit shelf: agent-ops desk vs AI-employee kits.
Soft next step
Name the Lead. Put never-send / publish / pay in its description. Prove one specialist handoff ends as a file the human can approve — not a send. Soft ladder: /free for the starter desk, Cheat Sheet for the roster/checklist mindset, then the Agent Starter Kit ($47) when you want the harness pack. No rush — ship the Lead gate into your own runbook either way.