Signed Client To Delivery Plan . Agency OS

What it does

Produces the full operational package for a new engagement: intake questionnaire, welcome email, project scope, client-facing proposal, milestone tracker, GHL sub-account setup guide, Conversation AI config, five workflow automations, milestone automation, and a handoff document. Every deliverable is written twice, as .md for your internal working copy and .docx for the client, into [target]/delivery/. It reads whatever ICA, offer, funnel, content and lead magnet work already exists in the workspace and builds on it instead of asking you to re-explain the client. If the client workspace does not exist yet, it creates client--[name] and seeds context.md.

Say this to start

This skill has no button. You start it by saying what you want. Any of these will do it:

/deliver-client

> I signed [client name]

> onboard [client name]

> new client

> build the delivery plan for [client]

> set up a GHL sub-account for this client

When to reach for it

just signed a client and have no delivery plan
onboarding a new client this week
need an intake questionnaire and a scope document
client asked what happens in the first 30 days
setting up a GHL sub-account for a client
handing a finished build back to the client

When NOT to use it

If you actually want	Use this instead
winning the client in the first place through outbound prospecting	`agency-sales-system`
converting a product's inbound leads on a sales call	`sales-conversion-engine`
keeping or saving a client after handoff	`client-retention`
building a standing referral programme off the back of the engagement	`referrals`
writing the Conversation AI bot prompts this plan calls for	`conversation-ai-architect`
building the funnel pages and sequences being delivered	`funnel-architect`

Before you start

What you need	Why
An offer defined for this product, saved as `offer-output.md`	the delivery plan anchors to what the client bought, so the skill stops and offers to run `/build-offer` if it is missing	Required
The engagement facts: services being delivered, engagement type (done-for-you, done-with-you, training, hybrid), target start date, deadline, engagement value and payment structure	these drive the scope document, the phase breakdown and the milestone-based payment triggers	Required
Client business name and industry	used to create the workspace and to shape intake questions, pipeline stages and the Conversation AI personality	Required
A working `.docx` generation setup (the skill uses the docx npm library)	clients never receive `.md` files, and the mandatory gate check will not complete without the `.docx` version	Required
A GoHighLevel MCP connection	without it the skill still produces every sub-account, pipeline, automation and Conversation AI configuration, but as written step-by-step instructions you implement manually	Optional
Existing workspace builds: ICA, offer, funnel, content plan, lead magnet, Meta lead and VSL outputs	each one found pre-populates part of the proposal, scope and GHL setup; missing ones are asked about inline rather than blocking the run	Optional
`verified-claims.md` populated with consented proof	the proposal and welcome email are put through a claim check, so any number or result without a registry entry is genericised or flagged as a proof gap	Optional
The `gws` command available for Google Drive upload	if present the `.docx` files are uploaded to Drive with conversion, if absent that step is skipped silently	Optional

How it runs

Workspace setupChecks brand/ for a workspace matching the client. If none exists it asks for business name, industry, website URL and social handles, then creates client--[slug] with a seeded context.md and sets it active. If _default is active for what is clearly a client engagement, it warns you first.
Client discoveryReads the existing schema artifacts, tells you what it found, then asks conversationally only for what is missing: services, engagement type, value and payment structure, start and completion dates, external dependencies, client-side contact, and whether they already have a CRM.
Intake and onboardingGenerates the intake questionnaire covering business foundation, brand assets, digital access credentials, business intelligence, service-specific questions and Conversation AI inputs, plus a kickoff call agenda and a welcome email for post-signature.
Project scope and proposalBuilds the internal scope document: engagement summary, deliverables table, explicit out-of-scope list, measurable success criteria, phase breakdown (Foundation days 1 to 5, Build 6 to 18, Launch 19 to 25, Optimise and handoff 26 to 30), dependencies and risks, and communication cadence. Then the client-facing proposal from your offer and ICA data.
GHL sub-account configurationWrites the sub-account creation steps, pipeline stages, contact fields and tags, the Conversation AI agent spec, five workflow automations (new lead welcome, appointment reminder, post-appointment follow-up, won deal onboarding, lost deal re-engagement) and milestone automation, all adapted to the client's industry and sales cycle.
Delivery trackingTurns the phase breakdown into a milestone tracker where every line has a date, an owner, a measurable completion criterion and a binary status. Adds the weekly cadence structure and a status report template.
Handoff planningProduces the handoff document covering what was built, the maintenance guide, performance baselines, ongoing support options and referral or testimonial timing, plus a table of upsell opportunities with estimated value and trigger conditions.
Gate check and saveRuns all six gate steps on every deliverable: auto-score and fix until it scores 9 or above, save the .docx, upload to Drive if available, update the dashboard, ask whether you want a presentation deck, and save the build overview. Then writes client-project.md and appends everything to assets.md.

What you get

Intake questionnaire (.docx) with sections A to F and a kickoff call agenda
Welcome email template for post-contract signing
Project scope document (.docx, 2,000 to 4,000 words) with out-of-scope items and measurable success criteria
Client-facing proposal (.docx, 1,500 to 2,500 words) built from your ICA and offer data
Milestone tracker (.docx table) with dates, owners and binary status
GHL sub-account setup guide covering CRM, funnels, AI Ads and Conversation AI
GHL Conversation AI configuration spec with flows, knowledge base and guardrails
Five GHL workflow automations plus milestone automation, in table form with timing, channel and content per step
Handoff document (.docx, 2,000 to 3,000 words) with system docs, maintenance guide and support options
Upsell opportunity table with estimated value and trigger condition per path
All of the above as both .md and .docx in [target]/delivery/
[target]/client-project.md schema artifact with status, phase and completion percentage
An updated agency-dashboard.html, a build-overview.md, and an optional cumulative agency-deck.pptx
A new client--[name] workspace with context.md if one did not exist

Honest limits

Read this before you rely on it

It writes the GHL configuration, it does not click through GHL for you unless the GoHighLevel MCP is connected. Without the MCP you get implementation-ready documentation to follow by hand.
It does not set your prices. It gives a floor of $997 for a one-time done-for-you build and $97 per month for the white-labelled sub-account, then uses your numbers when you supply them.
The proposal and welcome email pass through a claim check before you see them. Specific numbers, results and testimonials that do not trace to verified-claims.md come back generic or marked as a proof gap, so you need real proof on file to get specific copy.
The timeline it produces is a commitment you have to keep. It will push back on vague scope and over-committed dates, but it cannot deliver the work.
Demo mode skips the interview and fabricates content tagged [DEMO], with the file status set to demo rather than completed. That output is for showing the system, not for a real client.

Where people go wrong

The mistake	Do this instead
Running it before an offer exists for the product	Run `/build-offer` first, or accept the skill's prompt to chain into it. There is nothing to scope against otherwise.
Sending the client the `.md` files	The `.md` copy is your internal working copy and schema artifact. Send the `.docx`, a Google Doc, a `.pptx` or HTML.
Running the whole engagement out of the `_default` workspace	Let it create `client--[name]`. Client context, assets and learnings stay isolated per client.
Accepting milestones with no owner or no measurable criterion	Every milestone needs a specific date, a named owner and a criterion that is either met or not. Push back where the skill has not already.
Re-answering questions the workspace already holds	Let it read `ica-output.md`, `offer-output.md`, `funnel-output.md` and `content-plan.md` first. It only asks for genuine gaps.
Leaving out-of-scope items undefined to seem accommodating	The exclusions list is what protects the engagement. Fill it in properly at scope stage.

Worth knowing

Run the client's ICA, offer, funnel and content builds before this one. The delivery engine reads all of them and pre-populates the proposal value stack, the scope deliverables, the GHL pipeline stages and the Conversation AI personality, which is what turns a long onboarding interview into a short one.