# Dreamineering Dreamineering is an open user guide for AI-native ventures and experiments. It helps people and AI agents work together on what matters: coordinated work toward a human-owned beneficiary outcome within explicit consent, authority, risk, and stop boundaries. ## Authorship and Audience **Claim state**: Best-known-now based on current published content. AI agents wrote much of this site for agents and humans reading together. People keep purpose, consent, consequential judgment, and stop decisions. Wik Raines helms Dreamineering, which publishes this guide and holds copyright (established October 2021 as Mental Model). The first reader is an unfamiliar person working with their AI agent on a real decision. Agents retrieve the same purpose, permission, and evidence contract. For more context, see [About Dreamineering](https://mm.dreamineering.com/about/). ## Purpose - Dreamineering supplies public orientation, knowledge, recipes, and evidence discipline. - Stackmates is the bounded coordination substrate being developed and tested. - Ideas contains public dogfood venture experiments, not proof of value or a portfolio showcase. - Humans own purpose, consent, consequential commitments, accepted risk, stop decisions, and outcome judgment. - Agents may retrieve, compare, prepare, execute within authorised bounds, preserve evidence, and propose corrections. ## Start here - [Decision orientation loop](https://mm.dreamineering.com/purpose) — name what matters; use Values, Beliefs, or Control where needed. - [Decision Trace](https://mm.dreamineering.com/purpose/control/#control-card) — copy one record and retain its ID and accepted revision through each handoff. - [Authorise Bounded Work](https://mm.dreamineering.com/playbook/capabilities/business/governance/authorise-bounded-work/) — one governed start. - [Compare Outcome](https://mm.dreamineering.com/playbook/capabilities/business/analytics/compare-outcome-correct-method/) — append evidence after work or stop, then return a human-owned next decision. Drafts may contain unknowns. Missing decision-critical purpose, authority, evidence criteria, stop, or recovery means HOLD the affected action and return the exact gap and its owner. Never invent acceptance. Leave Review unfilled until action or stop; beneficiary effects may remain unknown afterward. Keep accepted intention, prediction, and mandate unchanged. A changed mandate needs a new draft revision and renewed human acceptance. A review does not renew authority. A valid schema record is not permission or verified evidence. - [Decision Trace v1 schema](https://mm.dreamineering.com/schemas/decision-trace.v1.json) - [Blank JSON trace](https://mm.dreamineering.com/examples/decision-trace/blank-draft.json) - [Hypothetical worked example](https://mm.dreamineering.com/playbook/capabilities/business/governance/authorise-bounded-work/#worked-example) - [Conformance scenarios](https://mm.dreamineering.com/examples/decision-trace/conformance.json) ## Operating Model `Purpose and Values → testable Beliefs → bounded Control → Playbook capability → coordinated action → venture experiment → evidence → learning` Concise loop: `Meaning → Capability → Coordination → Venture experiment → Evidence → Better next move`. Begin with a named beneficiary, preferred future, present reality, protected constraints, and human authority. Choose the smallest reversible action only after those are clear. Compare mature evidence with the expectation declared before action and use the variance to improve the better next move. ## Repository Map **Claim state**: Routes are candidate entry points based on current site structure. - [Public front door](https://mm.dreamineering.com/) - [About Dreamineering](https://mm.dreamineering.com/about/) — context on authorship, audience, and how the pieces fit. - [Purpose](https://mm.dreamineering.com/purpose/) — the decision orientation loop to clarify direction. - [Values](https://mm.dreamineering.com/purpose/values/) — choose what is good. - [Beliefs](https://mm.dreamineering.com/purpose/beliefs/) — test your map of reality. - [Control](https://mm.dreamineering.com/purpose/control/) — choose the lever that moves what matters. - [Open the Playbook](https://mm.dreamineering.com/playbook/) — choose the smallest fitting recipe. - [Inspect venture experiments](https://mm.dreamineering.com/ideas/) — compare evidence maturity, human authority, and next test. - [Business idea engine](https://mm.dreamineering.com/ideas/engine/) — inspect how ingredients support human-governed testing. - [Complete public route index](https://mm.dreamineering.com/sitemap.xml) - [Machine-readable knowledge card](https://mm.dreamineering.com/.well-known/agent.json) - [Playbook route manifest](https://mm.dreamineering.com/playbook-manifest.json) ## Developer Resources These are knowledge resources, not callable APIs or MCP tools. - [Dreamineering overview](https://mm.dreamineering.com/developers/dreamineering.md) — coordination job, route, boundaries, and authority. - [Agent discovery contract](https://mm.dreamineering.com/developers/agent-discovery.md) — knowledge-only classification and unverified DID. - [Knowledge navigation guide](https://mm.dreamineering.com/developers/knowledge-navigation.md) — choose Purpose, Playbook, Ideas, or sitemap. - [Evidence and authority guide](https://mm.dreamineering.com/developers/evidence-authority.md) — use before consequential action or proof claims. ## How To Use This Project 1. State the named beneficiary and job to be done. 2. Use Purpose and Values, testable Beliefs, and bounded Control to orient the work. 3. Choose the smallest fitting Playbook recipe; do not invent a callable endpoint or authority. 4. Preserve human consent, consequential judgment, accepted risk, stop decisions, and protected constraints. 5. Treat each Idea as a venture experiment rather than proof. 6. Return the observed outcome, variance, exceptions, beneficiary evidence, changed decision, and better next move. Traffic, clicks, activity, output, pipeline, and revenue are navigation or operating signals. They do not prove beneficiary improvement without the declared outcome and mature evidence.