Value Map · Internal · v0.1 · 2026-08-11
The structured lookup tool. Who we serve, what hurts, what we do about it, and what we can actually prove — connected end to end. For the readable strategy narrative, see Positioning.
This page has no login. Anyone with the URL can read it. noindex,nofollow keeps it out of ordinary search results but is not access control. Treat it as shareable with the team, not confidential.
The product changes monthly. A document rots silently; this file makes every claim addressable, taggable as sustaining or dynamic, and re-renderable. Edit the data, re-run the generator, the page and any future copy follow.
Source of truth: data/positioning-map.json. Regenerate with python3 scripts/build-positioning-map.py. Do not hand-edit this HTML — the next build overwrites it.
The stakeholder chain
Alex, 2026-08-10 team meeting — modeled on Shoney's framework: stakeholder -> title -> pain points -> value drivers -> differentiators -> features -> outcomes -> benefits -> stories
The value lens
Alex Hormozi $100M Offers — Value = (Dream Outcome x Perceived Likelihood) / (Time Delay x Effort & Sacrifice)
01 · The value equation
Value = (Dream Outcome x Perceived Likelihood of Achievement) / (Time Delay x Effort & Sacrifice)
Our position: strong
Our position: strongest available lever, currently underused
Our position: contested — our biggest honest weakness
Our position: strong, and the clearest wedge against DIY
02 · Sustaining vs dynamic
Alex, 2026-08-10: distinguish sustaining value props from dynamic pieces so marketing does not reset every time the product changes.
True regardless of which engine, model, or feature ships next. Safe to build durable marketing on. These should change roughly never.
You own the hardware, your data, and your agent configuration
Architectural, not featural. No upstream release changes it.
Model and engine neutrality — you are not betting your business on one vendor
Gets STRONGER as the field churns. Engine churn is our feature.
Setup, support, diagnostics and repair are part of the product
The service layer is the moat; competitors ship capability without anyone to call.
Human stays in control — approvals, permissions, visible actions
A posture, not a feature. Competitors are moving the opposite direction toward no-approval autonomy.
Business context and memory persist across sessions, tools and agents
The capability may be renamed or re-architected; the promise does not change.
Turns conversations into completed work inside tools you already use
Outcome-level, not implementation-level.
True today, likely different in 6 months. Usable in copy but must carry a date and be re-verified. Never build an evergreen campaign on these.
Current runner and engine versionschurn: high
Versions change constantly. Never lead customer copy with a version number.
Specific cloud and local model lineupchurn: very high
Model availability and rankings change frequently. Verify before technical copy.
Mac mini M4, 16GB / 256GBchurn: medium
Apple refresh cycle. 'A full Apple computer you own' is the sustaining version.
Named tools — Company Brain, Content Machine, Workflow Scoping, Scout Guidechurn: high
Several are unshipped or partially built. Public content may only discuss what ships today.
Price and bundle structureunresolved
BLOCKER: product-truth.md says $1,499; the 2026-08-10 meeting discussed $900 + accessories. Do not state a price until Alex confirms.
ProductiveBot+ is freechurn: medium
Explicitly 'free for the foreseeable future' — designed to become paid tiers. Do not promise permanence.
Channel and connector listchurn: medium
Grows. Safe to say 'the tools you already use'; risky to enumerate exhaustively.
03 · Stakeholder map
The full chain for each buyer. Where a link is missing, it is marked missing rather than filled in.
Proof status
Owner-operator, e-commerce
Pizza World / Ryan — $90K impact in first 3 months
Source: docs/dora-questions-answered.md Part 2
Agency owner or in-house media buyer
none yet — gap — No named agency win in our record
Source: docs/dora-questions-answered.md Part 2
Owner, roofing / construction / home services
roofing CRM — referenced in the 2026-07-08 founder interview; no named customer or number
Source: MEMORY.md ICP guidance; roofing CRM use case from 2026-07-08 interview
Gap: This is our stated core ICP and it has the weakest proof. No named trades customer with a number. Highest-value gap on this page.
Consultant / fractional operator
Jed — working prototype in ~45 min after $50K-$200K dev quotes
Source: docs/dora-questions-answered.md Part 2
Attorney, financial advisor, healthcare-adjacent
none — no named customer in this segment
Source: Dar UX persona research (Doug); qa-index-answer-bank.md Q18
Caution: NEVER say HIPAA compliant. Separate architecture from certification — see the Q&A bank Q18 for the approved framing.
Developer, technical founder, power user
Feb/Mar early adopter cohort — our first successful buyers were technical early adopters; noted in the 2026-08-10 meeting
Source: docs/dora-questions-answered.md Part 2; earliest customer cohort
Caution: This buyer converts today but is NOT the growth market. They can also evaluate us against ChatGPT desktop and Otto on specs — see the competitive landscape.
No stakeholder matches that proof status.
04 · Master feature catalog
The unit of effort this removes. Written as work the customer no longer does themselves. Framework: Jobs-To-Be-Done for the task, Value Proposition Canvas for pain → benefit, and Hormozi's Value Equation for why it matters.
Buyer
Proof
| Core feature | Subfeature | What it is / does | Customer job / unit of effort | Pain → benefit | Personas | Product state |
|---|---|---|---|---|---|---|
| Owned WorkstationA dedicated Apple Mac mini that belongs to the customer and runs their agents around the clock. | Dedicated always-on machine | Runs agent work continuously without tying up the customer's laptop. | Being the machine that stays awake so work can finish while they are gone. | PainWork stops the moment they close the laptop or leave the office.BenefitWork continues overnight, over weekends, and while travelling. | Solopreneur with a Store · Trades / Service Operator · Performance Marketer | Verified |
| Owned Workstation | Customer-owned hardware and data | Local files, memory, credentials and agent configuration live on hardware they bought. | Removing the need to trust a vendor's cloud with business data. | PainCancel a SaaS seat and the AI evaporates along with everything in it.BenefitUnplug test: unplug it and you still hold the machine, local data, memory and configuration. | Privacy-Bound Professional · Technical Builder · Former Executive Going Independent | VerifiedCloud-model prompts still go to the selected AI provider. Do not imply that every workflow is fully local. |
| Agent and Model LayerSupported agent runtimes and model choices, configured as a usable system rather than a pile of engines. | OpenClaw runtime | Provides an open agent engine for tools, channels, skills and recurring work. | Installing and maintaining the underlying agent runtime. | PainA capable open-source engine still arrives as a technical assembly project.BenefitA supported runtime that is ready for business workflows. | Technical Builder · Solopreneur with a Store · Trades / Service Operator | Verified |
| Agent and Model Layer | Hermes Agent runtime | Adds a second supported agent harness, including strong local-model and tool-use workflows. | Evaluating and wiring a second agent harness when it fits the work better. | PainOne agent engine is not best at every model or task.BenefitThe workstation can evolve without rebuilding the business around one vendor. | Technical Builder · Former Executive Going Independent · Privacy-Bound Professional | PartialRuntime lineup and default provisioning are dynamic. Verify the current shipping image before public use. |
| Agent and Model Layer | Model and provider choice | Supports hosted providers such as Anthropic and OpenAI plus other runtime-supported providers. | Choosing and changing the model used for a particular kind of work. | PainBetting the business on one provider that may reprice, weaken, or change policy.BenefitA better model can be adopted without abandoning the owned workstation and business context. | Technical Builder · Former Executive Going Independent · Privacy-Bound Professional | VerifiedNot interchangeable without configuration. Provider/model support changes and must be date-verified. |
| Agent and Model Layer | Local model support | Can run compatible open-weight models on-device for workflows configured to stay local. | Keeping selected sensitive work off third-party model infrastructure. | PainSome confidential work cannot be sent to a hosted model provider.BenefitA local-only path is available when the model and workflow fit the machine. | Privacy-Bound Professional · Technical Builder | PartialDo not imply every model or workflow runs well on the base 16GB machine, or that external tools cannot create network traffic. |
| Agent and Model Layer | AMS intelligent model router | Classifies request complexity and selects a configured model tier while logging routing decisions and estimated savings. | Choosing a capable-enough model for each request instead of paying for the strongest model every time. | PainUsing one premium model for everything wastes budget; using a weak model everywhere damages quality.BenefitRoutine work can use lighter models while complex work escalates to stronger ones. | Technical Builder · Former Executive Going Independent | PartialThe router is live in the internal OpenClaw AMS stack, but model names, cost assumptions and the customer packaging path are dynamic/incomplete. Do not publish savings without current logs. |
| Business MemoryDurable, searchable context so the agent knows the business instead of restarting every session. | TotalRecall persistent memory | Stores and retrieves decisions, documents and past conversations with source attribution. | Re-explaining the business, its rules, and prior decisions at the start of every task. | PainTelling the same tool the same context over and over.BenefitTell it once. The second time is faster than the first. | Solopreneur with a Store · Performance Marketer · Former Executive Going Independent · Trades / Service Operator | VerifiedMemory is stored locally, but retrieved chunks may be sent to the active cloud model. The deferred SecuritySuite pre-scan of recalled content is not built; do not claim end-to-end local processing. |
| Business Memory | Shared cross-app recall | Will let approved apps retrieve only the relevant durable context through a shared, scoped contract rather than copying a transcript. | Manually carrying the right context between tools, threads and sessions. | PainEach tool has to be re-briefed or gets too much context.BenefitWork can pick up with the right approved context, not a blind transcript dump. | Former Executive Going Independent · Technical Builder | RoadmapShared app-to-app retrieval is M2 work and is not yet verified end to end. |
| Business Memory | Company Brain | Structures the customer's business context so every task inherits it automatically. | Briefing the agent on the business before each piece of work. | PainGood output still requires a long setup prompt every time.BenefitTasks start already knowing the business. | Former Executive Going Independent · Solopreneur with a Store | RoadmapNOT SHIPPED. Internal reference only — never state as available. |
| Getting Work DoneThe agent performs multi-step work across real tools, not just answers questions. | Computer and browser use | Operates applications, files and websites the way a person would. | The clicking, copying, downloading and re-typing between systems. | PainWork lives in five tools that do not talk to each other.BenefitThe gap between tools stops being the customer's problem. | Performance Marketer · Solopreneur with a Store · Former Executive Going Independent | Verified |
| Getting Work Done | Scheduled and recurring work | Runs jobs on a cadence and reports results without being asked. | Remembering to do the recurring thing, then doing it. | PainReports, follow-ups and check-ins slip because nobody has time.BenefitThe recurring work simply happens; they review instead of produce. | Performance Marketer · Solopreneur with a Store | Verified |
| Getting Work Done | Multi-agent delegation | Delegates bounded parts of a larger job to multiple specialized agents where the installed runtime supports it. | Splitting a large job into parts and supervising each one. | PainOne assistant is a bottleneck on anything genuinely big.BenefitBigger jobs finish without more of the owner's attention. | Technical Builder · Former Executive Going Independent | PartialRuntime-dependent and not a universal customer workflow. Do not promise that agents automatically check or repair one another. |
| Getting Work Done | Content Machine | Turns a rough idea into a production-ready short-form content brief. | Translating an idea into a usable production brief. | PainContent dies between having the idea and producing it.BenefitIdeas reach production instead of the notes app. | Solopreneur with a Store · Performance Marketer | PartialDeployed but not verified end to end. Check status before any public claim. |
| Getting Work Done | Workflow Scoping | Turns a plain-language description of a routine job into a researched, actionable plan. | Working out what a workflow should even be before building it. | PainThey know the job is repetitive but cannot specify the automation.BenefitA real plan instead of a vague intention. | Solopreneur with a Store · Trades / Service Operator | RoadmapUNWIRED MOCKUP with known defects. Demo only — do not imply availability. |
| AccessReach the agent from wherever the customer already works. | Multi-channel messaging | Connects during setup to Slack, Telegram, Discord and the web portal; additional runtime-supported channels can be configured later. | Learning and opening yet another piece of software. | PainNot at a desk; cannot sit down and type into a new tool.BenefitMessage it like a person, from the truck or the phone. | Trades / Service Operator · Solopreneur with a Store · Performance Marketer | VerifiedOnly Slack, Telegram, Discord and the web portal are presented as out-of-box setup paths. Other channels require later configuration and support can change. |
| Business Systems and ConnectorsGoverned ways for agents to work with business applications without making raw credential access the product boundary. | Connector capability broker | Implements scoped capabilities, grants, policies, approvals, audit records, vault custody and revocation in front of provider adapters. | Wiring integrations safely and deciding which agent may do what. | PainConnecting a business tool too often means handing an agent a password and hoping.BenefitA real authority kernel exists for connected tools, without making raw credentials the product boundary. | Privacy-Bound Professional · Performance Marketer · Technical Builder | PartialThe broker core is real locally; connectors and the customer-facing control surface are still in progress. Do not imply a complete customer-shipping connector platform. |
| Business Systems and Connectors | Business-tool skills and adapters | Packages procedures and supported connectors for tools such as Google Workspace, GitHub and Slack. | Moving information and actions between the agent and the tools where work already lives. | PainThe workflow breaks at every tool boundary.BenefitThe agent can participate in real operating workflows instead of stopping at advice. | Performance Marketer · Solopreneur with a Store · Technical Builder | PartialExact adapters, authentication paths and end-to-end support vary. Verify each named integration before public use. |
| Security and ControlLayered checks and human control around agent actions. These are useful safeguards, not a claim of risk-free or fully contained operation. | Approval gates / human in the loop | Supports confirmation before configured external or sensitive actions. | Catching a mistake before it reaches a customer or a mailing list. | PainFear of an autonomous tool sending the wrong thing to the wrong person.BenefitThe customer has a chance to catch a consequential action before it leaves the business. | Solopreneur with a Store · Privacy-Bound Professional · Trades / Service Operator | PartialDo not claim universal enforcement across every tool, channel or connector. The broker's T2/T3 approval model is architecture, not a customer-shipping control plane. |
| Security and Control | PromptGuard | Scores direct and indirect prompt-injection, role-override, obfuscation and social-engineering patterns across multiple languages. | Recognizing when untrusted content is trying to change the agent’s instructions or authority. | PainAn agent reads email, files and web pages that may contain hostile instructions.BenefitSuspicious instructions can be flagged, logged, warned on or blocked according to configured policy. | Privacy-Bound Professional · Technical Builder | PartialThe package and detector ship, but enforcement depends on the active integration and configuration. HiveFence network features must be separately enabled and verified. |
| Security and Control | SkillGuard | Statically scans skill packages for credential theft, code injection, prompt manipulation, persistence, exfiltration and evasion, then produces a risk score and findings. | Reviewing an unfamiliar skill’s source and behavior before installation. | PainInstalling a community skill can silently widen the agent’s access or leak data.BenefitA machine-readable PASS / CAUTION / WARNING / FAIL assessment before trust is granted. | Technical Builder · Privacy-Bound Professional | VerifiedSkillGuard is a pre-install scanner, not a universal runtime permission gate. Do not describe it as intercepting every tool call. |
| Security and Control | SecuritySuite validation tools | Validates commands, URLs, file paths and external content for command injection, SSRF, traversal, exposed keys and exfiltration patterns. | Checking risky inputs before a shell command, web request or file operation executes. | PainA non-technical owner cannot manually security-review every action an agent proposes.BenefitDeterministic local checks are available before sensitive tool use. | Privacy-Bound Professional · Technical Builder | PartialThe tools ship, but the suite’s own SECURITY.md says automatic tool interception is not yet available and manual integration is required. Do not claim universal automatic enforcement. |
| Security and Control | PromptGuard secret and authority policies | Detects secret-extraction requests, credential-path harvesting, fake authority, approval-scope expansion and permission-bypass coaching. | Spotting when a message is trying to obtain secrets or stretch one approval into broader authority. | PainA convincing chat message can turn a helpful agent into a credential-disclosure path.BenefitSensitive requests can be refused or escalated instead of silently obeyed. | Privacy-Bound Professional · Technical Builder · Solopreneur with a Store | PartialPolicy and detection logic exist, but actual blocking/notification depends on configured integration. Never claim credentials cannot be exposed under all circumstances. |
| Security and Control | Security event logs and local audit trail | Records configured PromptGuard and SecuritySuite detections and security decisions in local logs for later review. | Reconstructing what was flagged, when, and why after a security event. | PainWithout an audit trail, the owner cannot tell whether guardrails fired or silently failed.BenefitSecurity decisions can be inspected rather than trusted as a black box. | Privacy-Bound Professional · Technical Builder | PartialLogging is configuration-dependent and not a centralized customer dashboard. Avoid promising complete coverage or real-time monitoring of every action. |
| Security and Control | AMS CommandGuard | Hooks into shell-command execution to classify, log and block configured dangerous command patterns. | Reviewing each proposed shell command for destructive or privilege-escalating behavior. | PainA general-purpose agent can issue a damaging command faster than a person can inspect it.BenefitHigh-risk command patterns can be stopped at the execution boundary. | Technical Builder · Privacy-Bound Professional | PartialCommandGuard is live in the internal AMS stack, but the customer Skill Store/package path is still in progress. Treat it as internal/limited, not universally shipped. |
| Security and Control | Local Runtime Secret Relay | Reads one explicitly allowlisted 1Password secret into a short-lived macOS Keychain cache and exposes it through an audited, fail-closed local command. | Giving a recurring local agent task access to one API key without copying it into prompts, .env files, logs or scheduler configuration. | PainScheduled jobs either hang on 1Password or solve the problem unsafely by storing plaintext secrets.BenefitA bounded, revocable credential path with stale/missing/corrupt caches failing closed. | Privacy-Bound Professional · Technical Builder | PartialCurrent live scope is the macOS Hermes + Fireflies profile. Setup is human-confirmed and the broader OpenClaw/customer profile is not yet shipped. |
| Setup and ActivationGetting from a sealed box to a working agent without technical skill. | Preconfigured on arrival | Ships with agent software, memory, security and channels already installed. | Assembling and configuring an agent stack from parts. | PainThe DIY route is dozens of hours and it breaks silently.BenefitStart with a working system, not a setup project. | Solopreneur with a Store · Trades / Service Operator · Former Executive Going Independent | Verified |
| Setup and Activation | Guided setup wizard | Walks through providers, channels and first configuration step by step. | Knowing what to configure, in what order, and when it is done. | PainFinishing a technical step and still not knowing if it worked.BenefitSetup that tells you where you are and what is next. | Solopreneur with a Store · Trades / Service Operator | VerifiedDo not attach a time claim until measured across shipped units. |
| Support and RecoverySomeone to ask, and something that repairs itself, when it breaks. | Scout | Built-in support agent for setup, troubleshooting and next-step guidance. | Diagnosing a problem and knowing what to do about it. | PainWhen it breaks at 9pm there is nobody to call.BenefitHelp that already knows the system it is helping with. | Solopreneur with a Store · Trades / Service Operator · Former Executive Going Independent | VerifiedEscalation can dead-end without human pickup — a known defect. Do not overclaim the human backstop. |
| Support and Recovery | Doctor diagnostics | Checks system health, surfaces reports and provides supported repair paths; some faults can be repaired automatically. | Noticing something is wrong before it becomes an outage. | PainNo way to tell whether the machine is actually healthy.BenefitProblems get found before the customer finds them. | Technical Builder · Trades / Service Operator | PartialDo not promise universal self-repair or rollback. ProductiveBot updates use forward repair and support diagnostics. |
| Support and Recovery | ProductiveBot+ membership | Included free: priority support, Skill Store access, training and community. | Finding trustworthy answers on how to actually use it. | PainCapable tool, no idea what to do with it on day three.BenefitA path from working machine to real business outcome. | Solopreneur with a Store · Trades / Service Operator | VerifiedFree 'for the foreseeable future' — never promise permanence, and do not promise a human response SLA. |
| ExtensibilityThe system gains capability after purchase instead of ageing. | Skill Store | Browse and install new capabilities; some free, some paid. | Sourcing and wiring up a new capability from scratch. | PainHardware bought today feels obsolete in six months.BenefitIt gets more useful over time, not less. | Technical Builder · Solopreneur with a Store | Verified |
| Extensibility | Custom skills | Customers and the community can build skills that fit existing workflows. | Bending the tool to the business instead of the reverse. | PainSoftware forces its own process onto the business.BenefitAutomation shaped like how they already work. | Technical Builder | Verified |
| Extensibility | Engine-aware skill installation | Installs the appropriate native adapter for OpenClaw, Hermes, or both when a skill supports them. | Learning each engine's internal skill format and installing the same capability twice. | PainA useful capability becomes another compatibility project.BenefitInstall the ProductiveBot capability; the system handles the engine-specific wrapper. | Technical Builder · Solopreneur with a Store | Verified |
| Lifecycle and ManagementActivation, updates and future visibility across the workstation's agent stack. | License activation | Binds a purchased unit to its purchase email and machine during setup. | Proving entitlement and establishing a supported device record. | PainNo reliable link between the physical system, the purchaser and support eligibility.BenefitThe machine enters setup with a verifiable ownership and support record. | Solopreneur with a Store · Trades / Service Operator · Technical Builder | VerifiedDo not claim recurring billing or a fully verified Shopify-to-license feed. Those are not current product facts. |
| Lifecycle and Management | ProductiveBot update path | Delivers ProductiveBot app improvements and guides supported forward upgrades. | Finding, interpreting and applying ProductiveBot software updates manually. | PainA working agent stack can decay as runtimes, models and integrations change.BenefitThe owned system can improve over time with a supported forward-upgrade path. | Solopreneur with a Store · Trades / Service Operator · Technical Builder | PartialThere is no rollback promise. Verify the exact currently supported update path before publishing. |
| Lifecycle and Management | Agent Management System | Would expose agent health, model visibility, permissions and workspace readiness in one control surface. | Inspecting every runtime, permission and agent separately to understand system state. | PainAs the stack grows, nobody can see what is running or what authority it has.BenefitOne place to understand and govern the agent workspace. | Technical Builder · Privacy-Bound Professional · Former Executive Going Independent | RoadmapNOT SHIPPED. Treat current designs and architecture as roadmap, not availability. |
| Lifecycle and Management | Organization / multi-Mini control | Would let an administrator share governed knowledge and permissions across location or employee workstations. | Keeping multiple locations or operators aligned without sharing one unrestricted agent environment. | PainA multi-location business needs shared context and central control without erasing local boundaries.BenefitConsistent company knowledge with per-location ownership and permissions. | Trades / Service Operator · Technical Builder | RoadmapFUTURE ARCHITECTURE ONLY. No customer-shipping organization control plane exists. |
| Lifecycle and Management | Agent roster and configuration audit | Internal CLI tools list agents, models, inherited configuration, channel bindings, skills and workspace health. | Inspecting multiple agent configuration files separately to understand who is active and how each is configured. | PainA multi-agent system becomes impossible to reason about when models, skills and bindings are scattered across files.BenefitOperators can inspect the current agent roster and configuration from one audit surface. | Technical Builder | PartialBuilt as internal CLI/dashboard scripts, not the customer-facing AMS interface. The unified visual shell remains unbuilt. |
| Lifecycle and Management | Configuration change detection and backups | Detects shared configuration changes, flags cross-agent impact and creates before/after backups for manual recovery. | Watching configuration drift and preserving a recoverable copy before one agent affects another. | PainA small config edit can silently take unrelated agents or plugins offline.BenefitUnexpected changes become visible and a prior configuration remains available for manual recovery. | Technical Builder | PartialInternal tooling exists; this is not a customer-facing real-time governance UI, and ProductiveBot application updates still have no promised rollback path. |
Nothing matches that combination — which is itself the answer: no capability of that proof grade serves that buyer.
The catalog is a synthesis, not a product promise. Verified means shipping and demonstrable; Partial means a capability exists with a meaningful limit or incomplete end-to-end proof; Roadmap means not customer-shipping.
docs/product-truth.mdmemory/2026-08-10.md — stakeholder framework and sustaining vs dynamic requirementmemory/use-case-requests.md — requested jobs, onboarding and customer painsmemory/2026-07-09.md — connector broker and future multi-Mini control architecturememory/2026-07-02.md — engine-aware Skill Store adaptersshopify-live/sections/pb-enhancements-page-detailed.liquid — current public feature catalogskills/prompt-guard/SKILL.md — installed detection modes and configurationskills/skillguard/SKILL.md + artifacts/release-1.0.6-review/manual-test-changelist.md — scan-before-install boundaryskills/security-suite/SKILL.md + artifacts/security/local-model-security-posture-review-2026-07-31.md — manual-validation versus automatic-hook boundaryskills/prompt-guard/SKILL.md — injection detection, secret/authority policies and known configuration surfaceskills/skillguard/SKILL.md + src/ — skill-package scanning and risk scoringskills/security-suite/SKILL.md + SECURITY.md — command/URL/path/content validation; manual-integration limitationskills/local-runtime-secret-relay/SKILL.md — allowlisted 1Password-to-Keychain relaydocs/agent-management-system-requirements.md — built CLI/config-guard components vs roadmap UIdocs/productivebot-system-map.md — current-state split between real broker, shipped engines and unbuilt AMS shell05 · Feature → so what
A feature with no "so what" does not belong in copy. Cautions are binding.
Buyer
Proof
| Feature | Kind | So what | Serves | Proof |
|---|---|---|---|---|
| Mac mini (owned hardware)Foundation | Sustaining | You own a real computer, not a seat in someone's cloud. Unplug test: cancel a SaaS and your AI evaporates; unplug a ProductiveBot and you still hold the whole system. | Solopreneur with a Store · Trades / Service Operator · Privacy-Bound Professional · Technical Builder | Verified |
| TotalRecall memoryMemory | Sustaining | Tell it once. It stops re-explaining your business every session — the single most underrated capability in customer conversations. | Solopreneur with a Store · Performance Marketer · Former Executive Going Independent · Trades / Service Operator | Verified |
| Scout (support + guidance)Support | Sustaining | Someone to ask when it breaks. ~98% resolution over 1,259 conversations.Escalation currently dead-ends without human pickup — a known defect. Do not overclaim the human backstop. | Solopreneur with a Store · Trades / Service Operator · Former Executive Going Independent | Verified |
| Doctor (diagnostics)Support | Sustaining | The machine can check and repair itself before you have to think about it. | Technical Builder · Trades / Service Operator | Verified |
| Guided setup wizardActivation | Sustaining | You are not handed a terminal. This is the difference between owning the parts and owning a working system.Do not attach a time claim until measured across shipped units. | Solopreneur with a Store · Trades / Service Operator · Former Executive Going Independent | Verified |
| PromptGuard / SkillGuard / SecuritySuiteSecurity | Sustaining | Layered prompt-injection detection, pre-install skill scanning and local validation tools — safeguards with documented enforcement limits, not a universal permission gate.SkillGuard is a pre-install scanner, SecuritySuite needs manual/optional integration for automatic interception, and PromptGuard depends on active configuration. See the Master Feature Catalog for the exact boundary. | Privacy-Bound Professional · Technical Builder | Partial |
| Connector brokerSecurity | Dynamic | The planned control plane gives agents scoped capabilities instead of raw credentials.Architecture and prototype work exist; this is not yet a customer-shipping connector control plane. | Privacy-Bound Professional · Performance Marketer | Roadmap |
| Model and engine neutralityFoundation | Sustaining | When a better engine ships — and one ships monthly — you are not stranded. Churn is our feature, not our risk. | Technical Builder · Former Executive Going Independent · Privacy-Bound Professional | Verified |
| Multi-channel messagingAccess | Dynamic | Message it from the truck, the phone, Slack. It works where you already talk. | Trades / Service Operator · Solopreneur with a Store · Performance Marketer | Verified |
| Scheduled / recurring workExecution | Sustaining | Work that happens without being asked is the difference between a chatbot and an employee. | Performance Marketer · Solopreneur with a Store | Verified |
| Approval gates / human in the loopControl | Sustaining | It shows you before it sends. The wrong-email scenario should not happen. | Solopreneur with a Store · Privacy-Bound Professional · Trades / Service Operator | Verified |
| Skill StoreExtensibility | Dynamic | The system gains capabilities after purchase instead of aging. | Technical Builder · Solopreneur with a Store | Verified |
| ProductiveBot+ membershipSupport | Dynamic | Included free today: support, Skill Store, training, community.Free 'for the foreseeable future' — never promise permanence. Do not promise a human response SLA. | Solopreneur with a Store · Trades / Service Operator | Verified |
| Local model supportFoundation | Dynamic | Removes third-party model traffic entirely for sensitive work. | Privacy-Bound Professional · Technical Builder | Verified |
| Company BrainMemory | Dynamic | Every task inherits the customer's business context.NOT SHIPPED. Public content may only discuss what ships today. | Former Executive Going Independent · Solopreneur with a Store | Roadmap |
| Content MachineExecution | Dynamic | A rough idea becomes a production-ready short-form brief.Deployed but not fully verified end-to-end. Check status before any public claim. | Solopreneur with a Store · Performance Marketer | Partial |
| Workflow ScopingActivation | Dynamic | Describe a routine job in plain language and get a real plan.UNWIRED MOCKUP with 3 known defects. Demo only. Do not imply availability. | Solopreneur with a Store · Trades / Service Operator | Roadmap |
Nothing matches that combination — which is itself the answer: no proof of that grade exists for that buyer.
06 · Awareness → hook type
Hormozi GOATed Ads: awareness level determines hook type. Source: memory/paid-media-foundation.md
| Awareness level | Hook type | Our status | Example |
|---|---|---|---|
| Completely Unaware | Curiosity-driven | Unused | — |
| Problem Aware | Pain-driven | Primary growth target | The busywork is not small. It is just scattered. |
| Solution Aware | Promise-driven | Active | The fastest way to get AI actually working in your business. |
| Product Aware | Proof-driven | Overweighted historically | OpenClaw and Hermes Agent, set up for real business workflows. |
| Most Aware | Offer-driven | Active | Ships ready to go on Apple hardware. |
The 2026-05 ad audit found most hooks were product-aware while the audience was solution-aware. That mismatch is the documented root cause of weak conversion — not merely 'ads do not work for this product.'
07 · Open questions
Listed rather than guessed. Each one blocks specific downstream work.
Larisa challenged the workstation framing on 2026-08-10; Alex received it well; unresolved.
Assembled from the Dora scenarios, Dar's personas and stated ICP guidance. Trades operator is our stated core ICP but has the weakest proof.
Agency and privacy-professional rows have no named win. Trades has a reference with no number.