Skip to content

guide.md

Latest commit

 

History

History
219 lines (159 loc) · 8.76 KB

guide.md

File metadata and controls

219 lines (159 loc) · 8.76 KB

Manifold for Non-Programming Decisions

A guide to using constraint-first thinking beyond software engineering.

Executive Summary

Manifold's constraint-first approach translates well to non-programming domains, with varying effectiveness:

Domain Applicability Best For
Research/Analysis HIGH Methodology design, study planning
Business HIGH Strategic decisions, expansion planning
Personal HIGH Major life decisions, career choices
Creative MODERATE Project planning (not creative direction)

Using the --domain=non-software Flag

To activate non-software mode, pass --domain=non-software when initializing a manifold:

/manifold:m0-init career-change --domain=non-software --outcome="Make the right career move"

This flag does two things:

  1. Switches constraint categories from software-specific (Business, Technical, UX, Security, Operational) to universal categories (Obligations, Desires, Resources, Risks, Dependencies)
  2. Switches artifact generation from code artifacts to decision artifacts (decision brief, scenario stress-tests, narrative guide, recovery playbook, risk watch list)

The flag defaults to software when omitted, so all existing workflows are unaffected. The reasoning mechanics (constraint types, tension analysis, backward reasoning) are identical in both modes -- only the vocabulary and output artifacts change.

Universal Categories

The five universal categories replace the software-centric categories when --domain=non-software is active:

Universal Category Core Question Replaces (Software)
Obligations What must/must-not be true? Business + Security
Desires What does success look like? UX + Business goals
Resources What can I bring to this? Technical (capability limits)
Risks What could break irreversibly? Security (broadened)
Dependencies What else must hold outside me? Operational

Constraint types (INVARIANT / GOAL / BOUNDARY) are unchanged. The categories change; the type taxonomy does not.

Key Findings

What Works Well

  1. Tension Analysis is Universally Valuable

    • Every domain tested produced meaningful tensions
    • Hidden dependencies emerged naturally
    • Trade-offs became explicit rather than implicit

  2. The "Risks" Category Surprises

    • Maps naturally to different concepts per domain:
      • Business: Financial/market risks
      • Personal: Exit strategies, safety nets
      • Research: Validity threats
    • Often surfaces considerations that emotional thinking overlooks

  3. Constraint Types Transfer Directly

    • Invariant (must never violate): Works in all domains
    • Goal (should optimize): Works in all domains
    • Boundary (hard limit): Works in all domains

What Needs Adaptation

  1. Creative Contexts

    • Framework better for planning than direction
    • Consider softer names: "Vision" instead of "Obligations"
    • Don't force structure on inherently fluid processes

  2. Time Expectations

    • 5-10 minutes is sufficient for most decisions
    • Complex decisions may warrant 30+ minutes
    • Don't over-engineer simple choices

Recommended Usage

Use Manifold When:

  • Decision has multiple stakeholders or competing interests
  • Stakes are high enough to warrant structured thinking
  • You suspect hidden constraints or trade-offs
  • Previous decisions in this area have gone poorly

Skip Manifold When:

  • Decision is simple or low-stakes
  • Creative exploration is the goal (not planning)
  • Time pressure makes structured analysis impractical
  • Intuition is likely sufficient

Quick Start for Non-Programmers

  1. Initialize with the non-software flag: 
    /manifold:m0-init my-decision --domain=non-software --outcome="Desired outcome"
  2. Brainstorm constraints across all 5 universal categories (2 min each)
  3. Identify tensions between conflicting constraints
  4. Resolve or accept each tension
  5. Decide with full constraint visibility

Category Deep Dive

Obligations (replaces Business + Security)

What must/must-not be true?

Ask yourself:

  • What are the non-negotiable requirements?
  • What legal, ethical, or contractual commitments apply?
  • What must absolutely happen -- or absolutely not happen?

Desires (replaces UX + Business goals)

What does success look like?

Ask yourself:

  • What does the ideal outcome feel like?
  • What would make stakeholders genuinely satisfied?
  • What experience is unacceptable?

Resources (replaces Technical)

What can I bring to this?

Ask yourself:

  • What time, money, skills, and energy are available?
  • What are the hard limits on what I can invest?
  • What capabilities exist today vs. what must be built?

Risks (replaces Security, broadened)

What could break irreversibly?

Ask yourself:

  • What's the worst-case scenario?
  • What's my exit strategy if this goes wrong?
  • What would I regret that I can't undo?

Dependencies (replaces Operational)

What else must hold outside me?

Ask yourself:

  • What external factors must remain stable?
  • Who else must act, and when?
  • What assumptions about the environment am I making?

Non-Software Artifacts

When --domain=non-software is active, m4-generate produces these artifacts instead of code:

Non-Software Artifact Software Equivalent Function
Decision brief Implementation code Decision with full constraint traceability
Scenario stress-tests Test suite Verifies decision holds under adversarial conditions
Narrative guide Documentation Reasoning survives time and personnel changes
Recovery playbook Runbooks Pre-decided responses to watch-list risks
Risk watch list + review triggers Dashboards + alerts What to monitor; when to reopen

See m4-generate-nonsoftware.md for full artifact templates.

Tension Types

Tensions surface naturally when constraints conflict:

Type Meaning Example
Trade-off Can't fully satisfy both Career growth vs family time
Resource Tension Not enough to meet both Speed vs quality with fixed budget
Hidden Dependency One requires the other first Customer experience requires local hiring

Command Applicability

Which Manifold commands work for non-programming decisions?

Command Applicable? Notes
/manifold:m0-init Yes Use --domain=non-software to activate universal categories
/manifold:m1-constrain Yes Discovers constraints across 5 universal categories
/manifold:m2-tension Yes Surface conflicts between constraints
/manifold:m3-anchor Yes Backward reasoning works for any outcome
/manifold:m4-generate Yes Generates decision artifacts (not code) in non-software mode
/manifold:m5-verify Yes Verifies completeness against constraints
/manifold:m6-integrate No Code-specific wiring
/manifold:m-status Yes Track decision progress

Key Insight: With the --domain=non-software flag, all phases through m5-verify are fully applicable. The artifact phases produce decision-appropriate outputs rather than code artifacts.

Privacy & Local Usage

For personal decisions containing sensitive information:

  • All manifolds are local - Files stay in .manifold/ on your machine
  • No data leaves your system - Framework operates entirely locally
  • You control the files - Delete anytime, no external dependencies
  • Private by default - No telemetry, no cloud sync, no sharing

Recommendations for sensitive decisions:

  1. Use generic/anonymized scenario names
  2. Keep manifold files outside version-controlled repositories
  3. Delete manifold files after decisions are made if desired

Templates

See scenario-template.md for a reusable template.

Example Scenarios

Conclusion

Manifold's constraint-first approach is not limited to software engineering. The core insight - that making ALL constraints visible BEFORE deciding leads to better outcomes - applies broadly to complex decisions.

The framework is most effective when:

  • Constraints exist across multiple categories
  • Trade-offs are not immediately obvious
  • The decision warrants structured analysis

For simpler decisions, the cognitive overhead may not be justified. But for the decisions that really matter, surfacing constraints and tensions before committing leads to fewer surprises later.