Overview

The Mission, Objectives, Milestones Clarification Framework — MOM — is the strategic-layer content producer of the Ora system. It is the framework that takes a raw idea, tension, or goal and turns it into the structured strategic hierarchy that drives every other piece of work: a Mission with Core Essence and Emotional Drivers, a classified set of Constraints, a list of Objectives, and the type-appropriate milestone structure (Active and Aspirational milestones for Projects, recurring milestones plus Aspirational maturity gates for Operations, Practices and Directions of Travel for Passions, a Critical Unknown plus exploration plan for Incubators). Where the Problem Evolution Framework supervises strategic-layer content, MOM produces it.

The framework’s first job is classification. Layer 1 runs four tests in order, first match wins. The Project Test asks whether the work has a specific, tangible deliverable with a finite endpoint at delivery. The Operation Test asks whether the work produces recurring deliverables on a cadence with no terminal endpoint until sunset criteria are met. The Incubator Test asks whether there is a central driving question the work is trying to answer (a Critical Unknown that will eventually resolve into one of the other classifications). If all three fail, the work is a Passion — ongoing exploration without required deliverables. The four classifications are mutually exclusive at the core; domain types (book, knowledge, workflow, fiction) compose with the core classification. When invoked under PEF supervision (M-Supervised mode), MOM may be passed a project_type from the calling context and dispatch directly without re-running the test; when invoked standalone (M-Standalone mode), MOM runs the test fresh.

The framework’s second job is endpoint discipline. For Projects and Incubators, the Resolution Statement Objectivity Protocol runs three checks against the candidate endpoint statement. Ambiguous Language Detection identifies words that allow multiple interpretations and forces resolution to one. Near-Miss Elicitation surfaces the patterns where the project could be declared complete without the underlying problem being solved — these become the matrix’s Excluded Outcomes, the runtime guards against silent substitution. Definition-Drift Detection identifies the failure modes where the Resolution Statement could quietly weaken across iterations. For Operations, the parallel Service Statement Objectivity Protocol runs the same three checks adapted to cycle-shape: a Cycle-Inspectability Check replaces Ambiguous Language Detection (does each cycle’s outcome admit a clear pass/fail at close?), Near-Miss Elicitation surfaces the cycle-shape near-miss patterns (cadence met but quality degraded; corpora consumed but unchanged; output produced but not consumed; maturity gate gamed not earned), and Service Statement Drift Detection guards against silent weakening of the cycle commitment.

The framework’s third job is constraints. Constraints are elicited as a byproduct of the Define and Analyze questioning rather than through a separate interrogation, then classified into three buckets. Hard Constraints are real, immovable, and non-negotiable for this matrix. Soft Constraints are preferences that should be honored but can be relaxed under specific conditions. Working Assumptions are things treated as constraints for now but with revisit triggers — they fire reconsideration when specified events occur. The classification matters because the three buckets behave differently downstream: Hard Constraints are Lock-protected; Soft Constraints can be relaxed with explicit Decision Log entries; Working Assumptions are revisited at trigger time and may convert to Hard or Soft once the trigger fires.

The framework’s fourth job is the type-appropriate milestone structure. Projects get an Active / Aspirational split: Active milestones are P-Feasibility-verified by the Process Inference Framework — there is a known process to deliver them; Aspirational milestones are not yet feasibility-verified and carry Contingency notes naming what would need to be true for them to become Active. The Promotion Protocol (in PEF) promotes Aspirational to Active as Contingencies resolve. Operations get recurring milestones plus maturity gates: each recurring milestone is the per-cycle commitment (Cycle Close Verification), and Aspirational maturity gates fire after extended operation (100 cycles shipped clean; quarterly external review passed for a year). Passions get Practices and Directions of Travel with no terminal milestones to verify. Incubators get a Critical Unknown plus exploration plan — the Critical Unknown is the matrix’s endpoint analog; resolution converts the Incubator into one of the other three classifications.

Two cross-cutting features matter in practice. The minimal-mode flag lets MOM elicit only the foundational fields appropriate to the classification (3–5 questions total) and accept “indefinite,” “none,” or “skip” for everything else. This is the Friction Principle in action — at the lower end of complexity, low-friction elicitation produces a usable matrix; over-eliciting produces an abandoned framework. The No-Punt Rule binds MOM’s M-Supervised escalations: if Layer 1’s test produces a different classification than PEF expected, MOM produces a structured reclassification report (reformulation as original type, pursue as recommended type, explore further) rather than silently demoting the work to a lighter classification.

Systemic context

MOM is the producer half of the Strategic Supervision System. [[Framework — Problem Evolution]] is the consumer/supervisor: PEF auto-invokes MOM in M-Supervised mode at PE-Init (to populate the strategic layer of a new matrix) and at PE-Iterate (when the strategic layer needs refresh). PEF v3.0’s MOM Invocation Protocol passes project_type and the Operation entry mode where applicable, and MOM dispatches directly without re-running the qualification test. MOM in M-Supervised mode invokes [[Framework — Process Inference]] in P-Feasibility mode for each Active milestone (Project) or recurring Active milestone (Operation), receiving a verdict that drives the Active/Aspirational split. Layer 5.5 creates the matrix file in Matrix/ per the type-appropriate template (sourced from [[Framework — Operations Manifest]] Appendix A for Operations; from MOM’s own Layer 5.5 for Projects, Passions, and Incubators). When Outcome 2 fires (terrain not yet mapped), MOM hands off to [[Framework — Terrain Mapping]] for Project-typed matrices or recommends spawning a precondition Project for Operation-typed matrices. When Outcome 3 fires (classification mismatch), MOM produces the reclassification report per the No-Punt Rule. The full picture sits in [[Paper — Strategic Supervision System]].

Ora’s output

Example question: I want to write a book on the philosophy of attention. How do I structure this as a project?

Ora’s response (running MOM in M-Standalone mode):

This is a clear case for the Project classification — the work has a specific tangible deliverable (a book) with a finite endpoint at delivery (the book ships). I’ll walk through what MOM does with the candidate idea, layer by layer.

Layer 1 — Classification. I’ll run the four tests in order.

The Project Test passes. “A book on the philosophy of attention” is a specific tangible deliverable; “ships when published” is a finite endpoint. I don’t need to run the Operation, Incubator, or Passion tests — first match wins. Classification: Project.

Layer 2 — Constraints elicitation. I’ll ask about constraints as a byproduct of getting the working definition clear, then classify what you tell me.

I need to ask: what’s the timeline you’re working with? What’s your constraint on length (a 60,000-word trade book is a different project from a 200,000-word academic monograph)? Are you committed to a specific publication path (traditional vs. self-publishing) or is that open? What’s your constraint on time-per-week? Do you have a co-author, an editor, or a deadline imposed by an external party?

Suppose you tell me: timeline is 18 months; length around 80,000 words; self-published is the default but open to traditional if a publisher is interested; 8 hours per week available; no co-author; no external deadline. The constraints classify as:

Hard: none stated. (If you’d said “must finish before May 2027 because of a speaking engagement,” that would be Hard.)

Soft: 18-month timeline; ~80,000 words; 8 hours per week. These are preferences — slipping any of them is possible with a Decision Log entry naming why.

Working Assumption: “Self-published is the default.” Revisit trigger: if a traditional publisher expresses interest before the manuscript is complete, this assumption fires reconsideration.

Layer 3 — Mission formulation with Resolution Statement Objectivity Protocol. This is where MOM does its load-bearing work. The candidate Resolution Statement is “I will write a book on the philosophy of attention.” I run the three checks.

Ambiguous Language Detection. “Write a book” is too loose. Does written mean drafted? Edited? Published? Does book mean PDF? Print? Both? Resolution: tighten to “The manuscript is complete in publishable form (final draft, edited, formatted for self-publication on Amazon KDP and ingested into the print-on-demand pipeline) and the book is live for purchase.” That sentence admits no interpretation under which it’s both true and not-true at the same time.

Near-Miss Elicitation. What would count as completing the project but not solving the underlying problem? Three candidates: (a) the manuscript is published but it’s a collection of loosely related essays rather than a coherent treatment of attention; (b) the manuscript is published but it covers neuroscience rather than philosophy and the title is misleading; (c) the manuscript is published but the philosophical position you wanted to develop never crystallized — the book is descriptive rather than argumentative. These three become the matrix’s Excluded Outcomes, recorded explicitly so PEF’s drift detection can fire against them at iterate time.

Definition-Drift Detection. What could quietly substitute for the Resolution Statement across iterations? Two candidates: “publish anything on attention” (drift toward generality) and “complete the manuscript as a draft I might publish later” (drift toward unfinishedness). The matrix’s Universal Problem-Definition Lock guards against both — changes to the Resolution Statement require explicit user-authorized PE-Iterate decisions.

The Mission’s Core Essence and Emotional Drivers go alongside. Core Essence: “Articulate a philosophical position on attention that synthesizes my reading and thinking into a coherent argument readers can engage with.” Emotional Drivers (in your voice): “I want to settle a question I’ve been carrying for years. I want to give the work a public form so it can be argued with. I want to have written this book because writing it is how I’ll think it through.”

Layer 4 — Objectives and milestone refinement. Objectives are strategic directions framed as continuous action. For this Project: “To articulate a coherent philosophical position on attention,” “To research and integrate the relevant literature on contemplative practice, cognitive science, and phenomenology,” “To draft and revise the manuscript to publication quality.”

Milestones split into Active and Aspirational. Active milestones must be P-Feasibility-verified (running PIF on each — is there a known process to deliver this milestone?). Today’s Active milestone for the manuscript: “Research phase complete — annotated bibliography of 40 primary sources with one-paragraph synthesis per source.” P-Feasibility: Reachable via standard scholarly research practice. Aspirational milestones today: “First draft complete (80k words),” “Manuscript edited to publishable quality,” “Published on Amazon KDP and live for purchase.” These get Contingency notes — what would need to be true for each to become Active. As prior milestones complete, the Promotion Protocol (in PEF) promotes them.

Layer 5.5 — Matrix file creation. I’ll create Matrix/Project Matrix Philosophy of Attention.md with project_type: project, the populated strategic layer (Mission, Resolution Statement, Excluded Outcomes, Constraints, Objectives, Active and Aspirational milestones), an empty Iteration History, and an empty Decision Log. The matrix is now ready for PE-Iterate at any time you have work to recap.

That is what MOM produces for a Project. The classification was easy; the load-bearing work was Layer 3 — the Resolution Statement Objectivity Protocol caught vagueness, made the near-miss patterns explicit, and named the drift modes the matrix will be supervised against. The matrix you have at the end is small (one screen of structured markdown) but tightly defined; the supervisory machinery (PEF’s drift check, Process Coherence’s runtime enforcement, the Promotion Protocol’s milestone promotion) all has the substrate it needs to do its job.

Commercial AI comparison

Comparison content auto-populates when the comparison-refresh framework runs against this question. Drafters do not author this section.

Brief comparison commentary

Auto-populates with the comparison content above.

How to use this framework

You can run the MOM pattern with any AI of your choice. The composition is single-pass for M-Standalone mode and two-pass when invoked under PEF supervision.

For M-Standalone, the prompt is:

[Paste the framework specification]

Run M-Standalone.

Idea: [Your description of the idea, tension, or goal in plain prose. One paragraph is usually enough.]

Constraints (optional): [Anything you already know — time, budget, scope, hard requirements.]

The AI returns: a classification (Project / Operation / Passion / Incubator) with rationale, a populated strategic-layer hierarchy in matrix-file format, the Objectivity Protocol report (for Projects, Operations, and Incubators), and an Active/Aspirational milestone split (Projects) or recurring + maturity-gate split (Operations) or Practices + Directions of Travel (Passions) or Critical Unknown plus exploration plan (Incubators).

For minimal-mode invocation (low-friction personal routines), add minimal: true to your prompt. The AI will elicit only 3–5 foundational questions and accept “skip” / “none” / “indefinite” for everything else.

For M-Supervised invocation (under PEF), the typical pattern is to paste both the PEF spec and the MOM spec, and let PEF dispatch — PE-Init or PE-Iterate auto-invokes MOM at the right moment with the right inputs. This is heavier but produces a fully PEF-supervised matrix from the start.

The framework is deliberately tool-agnostic. The four-test classification, the Objectivity Protocols, the Constraints classification, and the type-appropriate milestone structure are conceptual disciplines that survive the lift to any environment. The Active/Aspirational split is matrix-shape, not Ora-shape, and runs the same against any model.

One caution. The Objectivity Protocols are the load-bearing methodology — and they require honest answers. If you don’t articulate the near-miss patterns truthfully (Layer 3 Check 2), the Excluded Outcomes won’t catch silent substitution downstream. The framework can ask the questions; only you can answer them.

Other examples

  • A borderline classification case. “I want to start producing writing about leadership. What is this?” The Project Test asks for a tangible deliverable; “writing about leadership” doesn’t name one yet. The Operation Test asks for a recurring cadence; the user might or might not be committing to one. The Incubator Test asks whether there’s a Critical Unknown driving the work; the user might be exploring what specifically about leadership they want to say. The framework runs all three tests, surfaces what evidence each has, and offers the user a structured choice — the matrix is whichever classification fits the user’s actual commitment, not the one they default-named. Demonstrates the four-test classification in a borderline case where the user benefits from being asked which of the four they actually mean.

  • An Operation in minimal mode. A user wants to formalize a daily morning practice. MOM runs in M-Standalone with the minimal flag set. Three questions: what’s the practice (Service Statement)? What cadence (daily by 7am)? What near-miss pattern would worry you (you go through the motions but don’t actually settle into the practice)? The matrix is one screen, fits on one page, has Service Statement + Cadence + a single Excluded Outcome + an empty Performance Log. Total elicitation time: under five minutes. Demonstrates the Friction Principle — at low complexity, MOM produces small matrices that get used.

  • A reclassification under No-Punt. A user has been treating their newsletter as a Project (“ship 12 issues”) but the work has continued past issue 12 and they’ve started a new Project Matrix for the next 12. PEF iterate notices the pattern and invokes MOM for reclassification. MOM’s Layer 1 test classifies the work as an Operation (recurring deliverables on a cadence — the newsletter is a going concern). Outcome 3 fires. MOM produces the No-Punt reclassification report: option A (reformulate as a Project — “ship 50 more issues by Date X”), option B (pursue as an Operation — Service Statement “newsletter ships weekly indefinitely until sunset criteria are met”), option C (explore further — what’s the user actually committed to?). The user picks B. PEF v3.0 dispatches to OM-Convert for the Project-to-Operation conversion. Demonstrates the No-Punt Rule’s structured escalation in practice.

Citations

The Mission/Objectives/Milestones structure draws on goal-setting research in organizational psychology (Locke and Latham’s goal-setting theory) and the project-management distinction between strategic and tactical layers. The four-classification system — Project / Operation / Passion / Incubator as mutually exclusive core types — is internal to Ora and emerged from the cumulative design sessions that produced the Operations Manifest in 2026-05-08.

The Resolution Statement Objectivity Protocol’s three checks (Ambiguous Language Detection, Near-Miss Elicitation, Definition-Drift Detection) draw on requirements engineering literature (Sommerville’s distinction between functional and non-functional requirements) and red-teaming techniques for surfacing hidden assumptions. The Service Statement Objectivity Protocol is the cycle-shape adaptation introduced in v3.0 (2026-05-08) when the Operation classification landed.

The Constraints classification (Hard / Soft / Working Assumption with revisit triggers) draws on constraint-satisfaction frameworks in scheduling and the planning literature’s treatment of soft vs. hard preferences. The Active/Aspirational milestone split with P-Feasibility verification is internal to Ora and resolves a chronic problem in goal-setting where aspirational milestones masquerade as actionable ones.

The framework is single-author and originated 2025-09-04; v2.0 (2026-04-23) added M-Supervised mode and the Resolution Statement Objectivity Protocol; v3.0 (2026-05-08) added the four-classification dispatch, the Service Statement Objectivity Protocol, Operation entry modes, minimal-mode invocation, and the three-outcome branching with No-Punt reclassification.

Downloads

  • Framework specification (PDF) — link to ora-ai.org canonical artifact when published
  • Framework specification (plain text) — link to ora-ai.org canonical artifact when published
  • Full white paper (PDF) — link when published