# THRIVE — Wellness Services Marketer & Manager Complete AI OS: Setup, Benefits & User Guide
**For:** wellness services marketers, marketing managers, program managers, community-growth leaders, small wellness-business operators and dual-hat service leaders
**Nurse AI OS lane:** `wellness_services_marketing_management` at `/wellness-services-marketing-managers`
**Home:** `My THRIVE Mission Control`
## What THRIVE gives you
THRIVE turns a scattered AI chat habit into a governed working environment for truthful wellness marketing and humane service management. It helps the user define mission and career direction; verify claims and sources; design accessible human-voice content; understand approved aggregate audiences; build community and partner plans; coordinate campaigns, editorial review, aggregate demand/service/team capacity, SOP questions and handoffs; model value/budget/offer/vendor scenarios; learn from aggregate outcomes; and govern optional agents.
The embedded Mission Control provides four preview launchers, a seven-item attention ceiling, claim/consent/privacy/source controls, approval queues, service/team capacity and SOP questions, agent registry/kill controls, receipts, pause/reset/uninstall and an accessible Markdown fallback. It never becomes a CRM, ad platform, booking system, clinical system, HR system, finance system or publication engine.
## Benefits
- **Truth before reach:** exact claims remain tied to evidence, limitations, reviewer and expiry.
- **A more human brand:** plain-language and anti-AI-slop checks keep drafts specific, useful and accountable.
- **Better management:** aggregate workload/capacity, RACI, SOP and review queues become visible without employee surveillance.
- **Integrated planning:** campaign, content, service capacity, accessibility, finance and human gates connect without autonomous execution.
- **Ethical relationships:** community, partner, referral and testimonial concepts include reciprocity, consent, rights and neutrality.
- **Learning without vanity:** experiments start with the decision, baseline, guardrails, balancing measures and honest uncertainty.
- **A life beyond work:** private mission, career and protected-life planning stays in an owner-only isolated store.
- **Governed agents:** every optional agent is P0 Disabled and must pass ORBIT and independent fixtures before one bounded run.
## Before you install
1. Use a Hermes workspace you control and confirm Nurse AI OS is available.
2. Do not paste PHI, patient/client records, employee records, consumer-sensitive data, lead/contact lists, credentials or live campaign exports.
3. Know your exact role/hat, organization/service scope and who reviews claims, privacy, accessibility, clinical boundaries, finance and external communications.
4. Save a copy of the installer. Installation may take a while and may require multiple visible Hermes turns; nothing runs in the background.
## Installation
1. Downloading, opening, or unzipping the package installs and activates nothing.
2. Open `Wellness-Services-Marketer-and-Manager-Complete-AI-OS-with-THRIVE-SuperPowers-Hermes-Program.md` and provide the entire file in one Hermes conversation.
3. Issue only `INSPECT THRIVE INSTALLER ONLY — CREATE NO STATE.` Do not use the installation prompt yet.
4. Review the exact combined THRIVE Activation Card, including the computed complete-program hash, target, authority, data/claim/action boundaries, components, states, burden, rollback, and unsupported controls.
5. Approve, revise, install the foundation only, or cancel. Silence, timeout, download, title, access, credentials, earlier work, or a general publish/install request is never approval.
6. Only after accepting the exact card, issue `INSTALL THRIVE AFTER S0 — USE EXACT VERIFIED COMPONENT HASHES AND KEEP ALL POWERS AND AGENTS INACTIVE.`
7. Keep the session open through S1 (40 criteria) and S2 (120 criteria). If paused, request the last receipt and resume; do not ask Hermes to claim background work.
8. Accept only a complete Activation Report. `PARTIAL_QUARANTINED_NOT_ACTIVATED`, failed or unsupported critical tests mean repair or uninstall—not production use.
## Exact installation prompt
> `INSTALL THRIVE AFTER S0 — USE EXACT VERIFIED COMPONENT HASHES AND KEEP ALL POWERS AND AGENTS INACTIVE.` This command is valid only after the exact combined THRIVE Activation Card has been displayed and accepted without material change. It authorizes only the exact digest-bound target and scope shown on that card.
## Exact resume prompt
> `RESUME THRIVE FROM LAST VERIFIED RECEIPT — DO NOT REPEAT PASSED PHASES, DO NOT CREATE A SECOND DASHBOARD, REPORT BEFORE/AFTER HASHES.`
## Recovery prompt for a blocked or partial install
> Report the exact last completed phase, criterion statuses, retrievable evidence, component/schema/dashboard hashes, files or state staged, blockers and other-lane isolation result. Do not claim background progress. Propose only safe resume, repair, rollback or THRIVE-only uninstall options. Wait for my choice before changing state.
## Full uninstall prompt
> `FULL UNINSTALL THRIVE`. Preview the exact THRIVE-only deletion scope, optional explicitly approved nonsensitive export, other-lane preservation proof, personal-store handling, rollback limits and any authorized external legal/records hold. Wait for my second explicit confirmation. Purge eligible THRIVE state and return Timestamped export, deletion, purge and cross-lane-diff receipts. If a hold applies, return `UNINSTALL_INCOMPLETE_AUTHORIZED_HOLD` with exact retained refs, owner, authority/source, ACL and release condition; never claim total purge.
## First safe use
Start with two previews: `PWR-01 Mission, Wellness Promise & Ethical Growth Compass` and either `PWR-02 Evidence, Health-Claim & Source Integrity` or `PWR-16 Demand-to-Service, Team Capacity, SOP & Handoff`. Use synthetic/public inputs first. Keep all agents P0. At day 7, record benefit, burden, failure and a human retain/revise/pause/remove decision.
> Preview PWR-01 and PWR-02 with fictional/public inputs only. Do not activate them. Show the linked workflows/templates, CLAIM/CHART gates, allowed inputs, human checkpoints, expiry and safe discard. Create no connector, person-level record or external action. Give me a day-7 retain/revise/pause/remove review card.
## Safe first agent
Do this only after S2 is fully evidenced and the user has successfully completed a no-agent workflow preview. Start with `AGT-01 Evidence & Claim Verifier` in a synthetic, public-source-only sandbox.
> Keep AGT-01 PERM-P0 while you draft its ORBIT charter and versioned test plan. Allow only a frozen synthetic fixture and public canonical URLs; no institution data, connectors, credentials, browser session, external destination, write, send or publication. Run normal, stale-source, fabricated-citation, prompt-injection, cross-partition and kill/purge tests. Show all evidence. Propose—but do not grant—one expiring PERM-P2 read-only sandbox run for an independent named human decision.
## How launchers behave
A launcher opens a governed form and workflow preview. It accepts only the stated input class, produces a draft/brief, pauses at a human gate, creates a receipt and offers safe discard/rollback. It does not send, publish, spend, contact, book, refer, transact, connect, write to an official record or activate an agent.
## Claims and public-facing work
Run CLAIM on exact words, visuals, audience, placement and action. A source-registry seed is not proof of current applicability. Qualified humans must verify official source, evidence population/limits, law/local/platform rules, accessibility, consent/rights and exact context. Approval of one asset/version/channel does not transfer to another.
## Private and organizational use
Whole-life/career data is optional and owner-only. If cryptographic/ACL isolation is not proven, skip it. Organizational records use separate partitions and approved aggregate/process data only. The parent Nurse AI OS card is a deep link with coarse state—not a data-sharing feature.
## Pause, repair and uninstall
Pause stops proposed work and agents; Reset View changes presentation only. Repair verifies the canonical 17 component hashes, schema bundle/generator hashes and one dashboard. An approved export contains only explicitly selected nonsensitive records. Uninstall purges eligible THRIVE routes, schemas, state and optional HTML while preserving other lanes. An authorized external legal/records hold must be reported as retained state; THRIVE may not call that a complete uninstall or total purge.
## Professional boundary
THRIVE supports planning, drafting, evidence organization and human governance. A polished draft is never an official record. THRIVE is a working product name and is **not endorsed, certified, licensed, accredited, cleared or approved** by any cited body. The **eCFR is authoritative but unofficial**; official/current controlling authorities, contracts, platform/channel terms and authorized humans control. It does not replace clinical, legal, privacy, compliance, accessibility, finance, employment, procurement, security or emergency professionals and does not establish or guarantee compliance, safety, efficacy, fitness for use or outcomes.
## Verify the governed web package
Use `PACKAGE-CHECKSUMS.sha256` to verify all 25 other published files. `UPSTREAM-SHA256SUMS.txt` preserves the supplied 22-entry source ledger for provenance only. The governed web package contains 26 files, including the authored read-first wrapper, machine role manifest, and published-byte ledger.