← All runbooks
gooseworks-ai / playbooks-client-onboarding

Client Onboarding

Run the full onboarding playbook for a new client. This runbook orchestrates three phases of work: intelligence gathering (parallel deep research across company, competitors, founders, SEO, AEO, paid ads, industry, and GTM), synthesis of all research into a Client Intelligence Pa

agent claude-codemodel claude-sonnet-4-6snapshot python312-uveval programmatic6 stepsv1.0.0

Deploy Client Onboarding to your jetty.io

One-click installs this runbook into a collection on your Jetty account. You can run it from the Spot dashboard, schedule it, or pipe inputs in via the API.

Deploy on jetty.io →View source
The shape of the run

6 steps · start to finish.

  1. 1
    Step 1

    Environment Setup

    ▶

    Validate inputs and create the client folder structure before beginning any research.

    echo "=== CLIENT ONBOARDING SETUP ==="

# Required inputs
if [ -z "${COMPANY_NAME}" ]; then echo "ERROR: COMPANY_NAME is not set"; exit 1; fi
if [ -z "${COMPANY_URL}" ]; then echo "ERROR: COMPANY_URL is not set"; exit 1; fi

CLIENT_SLUG=$(echo "$COMPANY_NAME" | tr '[:upper:]' '[:lower:]' | tr ' ' '-' | tr -cd '[:alnum:]-')
CLIENT_DIR="clients/${CLIENT_SLUG}"

mkdir -p "${CLIENT_DIR}/intelligence"
mkdir -p "${CLIENT_DIR}/strategies"
mkdir -p "${CLIENT_DIR}/campaigns"
mkdir -p "${CLIENT_DIR}/leads"
mkdir -p "${CLIENT_DIR}/content"

# Initialize running notes
echo "# ${COMPANY_NAME} — Notes" > "${CLIENT_DIR}/notes.md"
echo "Onboarding started: $(date -u +%Y-%m-%dT%H:%M:%SZ)" >> "${CLIENT_DIR}/notes.md"
echo "URL: ${COMPANY_URL}" >> "${CLIENT_DIR}/notes.md"

mkdir -p /app/results
echo "Setup complete. Client dir: ${CLIENT_DIR}"

    Expected folder structure after setup:

    clients/<client-name>/
├── context.md              # Populated during research
├── notes.md                # Running log (initialized above)
├── intelligence/           # Phase 1 outputs
├── strategies/
├── campaigns/
├── leads/
└── content/
  2. 2
    Step 2

    Phase 1 — Intelligence Gathering (max 8 parallel tasks)

    ▶

    Run all eight research tasks. Groups A and B may run in parallel with each other; tasks within each group may also run in parallel.

  3. 3
    Step 3

    Phase 2 — Synthesis & Diagnosis

    ▶

    > **Human Checkpoint**: Review the Intelligence Package for accuracy before generating strategies.

  4. 4
    Step 4

    Phase 3 — Strategy Generation (max 3 rounds)

    ▶

    > **Human Checkpoint**: Review strategies with client before implementation.

  5. 5
    Step 5

    Iterate on Errors (max 3 rounds)

    ▶

    If any Phase 1 research output is missing or incomplete:

  6. 6
    Step 6

    Final Checklist

    ▶

    echo "=== FINAL OUTPUT VERIFICATION ===" CLIENT_DIR="clients/${CLIENT_SLUG}"