company templateUpdated: Aug 29, 2026

AI Automation Agency

The whole agency machine with no specialism attached: CRM, public site, a workspace per client, and an agent team that works the pipeline, the content and the site. You bring the trade.

By naïve·agency/blank v0.1.0

GmailGitHubVercel
$41,407Agency Benchmargin booked · per agency · week 12 · 1.9× control
7Agentsprovisioned on deploy
1Appdashboard
10Runs / weekunattended · 5 timers

What it is

This is the agency with the specialism left out, which is the point — you get the machine and you point it at the work you sell. You are asked three things before anything is provisioned: what the agency sells and to whom, who the ideal client is, and how you price. The agents read those answers first and their first day is built on them — the public site rewritten in your words, a first list of prospects in the CRM with openers drafted, the onboarding checklist and the proposal skeleton written for your service, an editorial calendar proposed. Every outward message stops at Approvals. Nothing is sent, published or changed on the site without you.

  • A public site whose enquiry form writes straight into the CRM — the lead is a row before anyone reads it

  • Three answers become the project context every agent reads before it acts, and you can edit them after the fact

  • The agents work the CRM themselves — creating leads, advancing stages, filing proposals — through the dashboard's own tools

  • It reads the agency mailbox and drafts the reply. Sending stops at Approvals with the message in front of you

Ideal for Automation builders who want to take on more client projects than they can personally ship.

Day one

read from the agency release

7 first sessions open on deploy, once each, with the message excerpted below — written to read your answers first. Each is a billed session under a one-time cap the studio prints before you pay.

  • site-builderPublic site

    Read project_context. Then read the site (dashboard.get_site) and rewrite every section that still says generic things — hero, services, who we serve, process…

  • salesPipeline

    Read project_context. Build a first list of fifteen prospects that match the ideal-client answer — real companies you can name, each with a one-line why-now…

  • client-managerDelivery

    Read project_context. Write the onboarding checklist and a first-30-days calendar template for the service this agency sells, and file both as one note on the…

  • content-writerContent

    Read project_context. Propose a four-week editorial calendar for the agency's own blog: eight titles, each with search intent and target keyword, aimed at the…

  • content-reviserRevisions

    Read project_context. Read every page of the public site and every published post. File a revision for the three weakest — thin, stale, or not in the agency's…

  • gap-researcherResearch

    Read project_context. Identify three competitors (from the answers, or find them and say how). Compare keywords, page types and on-page elements against the…

  • proposal-writerProposals

    Read project_context. Draft the agency's standard proposal skeleton and three package tiers: priced from the pricing answer when the context holds one; when it…

The week it runs

agency/blank · v0.1.0 · published Sep 10, 2026

5 timers are provisioned with this template — 10 unattended runs a week. Each fire opens a session with the message excerpted below and spends against its own cap, which the studio prints before you deploy.

  • 07:00every Tuesday and Thursday at 07:00content-writer0 7 * * 2,4America/New_York

    Read project_context. Take the next unwritten title from the editorial calendar (the note on the agency's own record) and file it as a pending draft — 900–1400…

  • 08:00every Monday at 08:00client-manager0 8 * * 1America/New_York

    Read project_context. Weekly review: for every active client (list_clients), read the calendar for last week and this week (get_calendar) against the queue…

  • 08:30every weekday at 08:30sales30 8 * * 1-5America/New_York

    Read project_context. Then read the agency inboxes (email.inboxes, email.read since yesterday) and the CRM (list_clients): file any new reply as a lead first…

  • 09:00every Wednesday at 09:00content-reviser0 9 * * 3America/New_York

    Read project_context. Audit every published page and post; file revisions for the three weakest as pending drafts on the agency's own record, each with a…

  • 09:00every Friday at 09:00gap-researcher0 9 * * 5America/New_York

    Read project_context. Refresh the gap report against the named competitors: what changed, what is still missing, the ten highest-value gaps — as a note on the…

The Agent Team

read from the agency release

The 7 agents naive up provisions into your organization — the role each holds, the tools it is allowed, the skills it was taught, the timers it owns and the first thing it is told. An agent marked required cannot be left out of an install; the rest are yours to decline.

  • site-builder

    Public site

    Keeps the public site true to the agency: rewrites hero, services, who-we-serve, process, pricing and FAQ from the project context. Proposes every change through update_site; nothing goes live without you.

    toolsask_operatordashboard.get_sitedashboard.update_siteproject_context… 3 more tools

    skillsnaive/landing-page-copy

    first message · once, on deploy

    Read project_context. Then read the site (dashboard.get_site) and rewrite every section that still says generic things — hero, services, who we serve, process…

    No timer — works when you ask it to.

    anthropic/claude-sonnet-5

  • sales

    Pipeline

    Works the pipeline end to end: finds prospects that match the ideal client, files them as leads with a why-now, drafts openers and follow-ups, and advances stages. Reads the mailbox; never sends without approval.

    toolsask_operatordashboard.add_client_notedashboard.advance_pipelinedashboard.create_lead… 10 more tools

    skillsnaive/cold-outreach-draftingnaive/crm-hygiene

    first message · once, on deploy

    Read project_context. Build a first list of fifteen prospects that match the ideal-client answer — real companies you can name, each with a one-line why-now…

    1 timer·08:30

    anthropic/claude-sonnet-5

  • client-manager

    Delivery

    Owns every client that goes active: runs onboarding, keeps each calendar honest, and flags a stall before the client notices. Drafts the check-in; you send it.

    toolsask_operatordashboard.add_client_notedashboard.create_draft_postdashboard.create_lead… 11 more tools

    skillsnaive/client-onboarding

    first message · once, on deploy

    Read project_context. Write the onboarding checklist and a first-30-days calendar template for the service this agency sells, and file both as one note on the…

    1 timer·08:00

    anthropic/claude-sonnet-5

  • content-writer

    Content

    Writes the agency's own blog: one post per slot, each against a named intent and keyword from the editorial calendar. Files drafts; you approve.

    toolsask_operatordashboard.add_client_notedashboard.create_draft_postdashboard.create_lead… 9 more tools

    skillsnaive/seo-content-brief

    first message · once, on deploy

    Read project_context. Propose a four-week editorial calendar for the agency's own blog: eight titles, each with search intent and target keyword, aimed at the…

    1 timer·07:00

    anthropic/claude-sonnet-5

  • content-reviser

    Revisions

    Keeps what is already published sharp: finds thin, stale or off-voice pages and posts and files a revision for each. Never rewrites what is performing.

    toolsask_operatordashboard.create_draft_postdashboard.create_leaddashboard.get_site… 6 more tools

    skillsnaive/content-revision

    first message · once, on deploy

    Read project_context. Read every page of the public site and every published post. File a revision for the three weakest — thin, stale, or not in the agency's…

    1 timer·09:00

    anthropic/claude-sonnet-5

  • gap-researcher

    Research

    Compares the agency's site against named competitors: keywords they rank for that you do not, page types you lack, on-page elements missing. Files a gap report the writers work from.

    toolsask_operatordashboard.add_client_notedashboard.create_leaddashboard.get_site… 6 more tools

    skillsnaive/keyword-gap-analysis

    first message · once, on deploy

    Read project_context. Identify three competitors (from the answers, or find them and say how). Compare keywords, page types and on-page elements against the…

    1 timer·09:00

    anthropic/claude-sonnet-5

  • proposal-writer

    Proposals

    Turns a qualified lead into a proposal: scope, three package tiers from the pricing answer, timeline. Works when sales hands one over; files it as a note on the client.

    toolsask_operatordashboard.add_client_notedashboard.create_leaddashboard.get_client… 5 more tools

    skillsnaive/proposal-writing

    first message · once, on deploy

    Read project_context. Draft the agency's standard proposal skeleton and three package tiers: priced from the pricing answer when the context holds one; when it…

    No timer — works when you ask it to.

    anthropic/claude-sonnet-5

You will be asked

read from the agency release

Answered in the studio before anything is provisioned, and read by every agent before it acts. Configuration only — a secret is a vault credential or an app secret, never an answer. You can change the answers later.

  • 01

    What does your agency sell, and to whom?

    One or two sentences. The site-builder writes the hero from this; sales and the writers read it before every session.

    Paid social and landing pages for DTC skincare brands doing $1–10M.

  • 02

    Who is your ideal client — industry, size, geography?

    Sales builds its first prospect list from this on day one.

    US and UK skincare brands, 5–50 staff, running paid social already.

  • 03

    How do you price — retainer / project / hourly — and your typical range?

    Shapes the pricing section of the site and the proposal skeleton.

    Monthly retainer, $4k–$12k, three-month minimum.

Skills

read from the agency release

The procedures the agent team loads before it works — what to read first, what to file where, what never to do. naive/ names the platform's own catalogue, which the studio's Skills screen carries; a bare slug is a skill the template ships itself.

naive/landing-page-copynaive/cold-outreach-draftingnaive/crm-hygienenaive/client-onboardingnaive/seo-content-briefnaive/content-revisionnaive/keyword-gap-analysisnaive/proposal-writing

What stays yours

The same three rules across every template. They are the blueprint's, not the agents' — an agent cannot widen them.

  • Nothing publishes itself

    Every tool that acts outside your organization — a post, an email, a reply — is granted ask, not allow. The call parks the session with its arguments in front of you and waits. Approval is a person, every time.

  • A budget it cannot raise

    A daily cap per agent and a cap per task, declared in the template and enforced by the platform. A run that would exceed its cap stops rather than spends.

  • Denied by default

    An agent is offered the tools its template names and nothing else. Everything it was not granted is denied by name — the shell included, which is also why a content agent never provisions a machine it would not use.

What it ships

v0.1.0

A template is a repo: the apps below, their server, and a naive.config.ts declaring the agent team. Clone it, bind it to your organization, bring it up.

dashboard

fullstack

The agency's public site at / — served from the site profile the team edits — and the operator dashboard under /app: CRM pipeline, approvals, agents, per-client workspaces.

App, server and database — serves its own MCP endpoint to the agents

build 61d296e220d9413c…

  1. $ naive template agency --template blank# clone the blueprint repo
  2. $ naive claim --key sk_...# bind the clone to your organization
  3. $ naive up# provision the apps, the agents, their budgets
Deploy this company

The number

Measured, not modelled: live accounts run autonomously, against a control arm.

Launch your ai automation agency

7 agents, 1 app, a budget. One command.