organon

--- name: organon description: >- This skill should be used when the user faces an engineering decision, asks "should we...", "which approach...", "trade-off between...", "how to structure...", "evaluate this design", "review this code philosophically", "think through this", "analyze this approach", "what principles apply here", or invokes /organon, /organon:decide, /organon:review. Also auto-triggers on architectural decisions, debugging sessions, code reviews, design evaluations, and any situation where structured philosophical reasoning improves the outcome. This skill turns 62 philosophical principles from 20 philosophers into a live reasoning engine for engineering. version: 1.0.0 --- # Organon — Philosophical Reasoning Engine > Source: [Organon](https://gitlab.com/lightcyphers-open/organon) by Lightcyphers SRL, licensed under [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/). A decision engine and review framework that applies 62 principles from 20 philosophers to engineering problems. Not a reference — an engine. Detect the situation, select principles, run the protocol, produce structured output. ## Modes **Auto-detect** (default): Read context, determine if this is a decision or a review, select depth, execute. **Decision mode** (`/organon:decide`): Engineering decision analysis. Takes a situation, identifies applicable principles, runs protocol steps, produces recommendation with objections addressed. **Review mode** (`/organon:review`): Philosophical code/design review. Takes code or design, systematically evaluates against quality dimensions derived from the 62 principles. ## Depth Levels | Level | When | Context loaded | Output | |-------|------|---------------|--------| | **Quick** | Clear situation, single principle applies | This SKILL.md only | 3-5 lines: principle + action | | **Standard** | Non-trivial, multiple principles apply | + relevant philosopher references | Principles with rationale + protocol summary | | **Deep** | Architectural, irreversible, high-stakes | Full protocol via philosopher-council agent | 22-step walk + Summa objections + confidence | Auto-select depth by: Quick if one principle clearly maps. Standard if 2+ principles or ambiguity. Deep if irreversible decision, architectural choice, or conflicting principles. User can override with `/organon quick`, `/organon standard`, `/organon deep`. ## Execution Protocol ### Step 1: Detect Situation Read the current context — what is the user doing? Map to one of these categories: | Context signal | Mode | Typical depth | |---|---|---| | "should we", "which approach", "trade-off" | Decision | Standard | | "review this", "what's wrong with", code snippet | Review | Standard | | "debug", "why does this", "root cause" | Decision (abductive) | Standard | | "architecture", "design system", "how to structure" | Decision | Deep | | Routine question, clear answer exists | Decision | Quick | | Explicit `/organon:decide` or `/organon:review` | As specified | As specified or auto | ### Step 2: Route to Principles Use the compact routing table below. Match the situation to applicable principles, then load the relevant reference files for those principles. ### Step 3: Apply and Produce Output **Decision mode output format:** ``` ORGANON — Decision Analysis =========================== Situation: [detected/stated] Depth: Quick | Standard | Deep Applicable Principles: #XX Principle (Philosopher) -> Concrete action for this situation #XX Principle (Philosopher) -> Concrete action for this situation [Standard/Deep only:] Protocol Steps: 1. Four Causes -> Material: X | Formal: Y | Efficient: Z | Final: W 2. Effectual Truth -> Verified: [what was checked] ... [Deep only:] Summa Method — Objections: Videtur quod non 1: [objection] -> Ad 1: [resolution] Videtur quod non 2: [objection] -> Ad 2: [resolution] Decision: [recommendation] Confidence: [High|Medium|Low] — Cumulative case: [N signals listed] ``` **Review mode output format:** ``` ORGANON — Philosophical Review =============================== Target: [file/code/design] Quality Dimensions: Kalokagathia (beauty=good): [pass|warn|fail] [finding] Golden Mean (balance): [pass|warn|fail] [finding] Structural Stability: [pass|warn|fail] [finding] Falsifiability (tests): [pass|warn|fail] [finding] Proportionate Causality: [pass|warn|fail] [finding] Beetle in the Box (contracts): [pass|warn|fail] [finding] Information Theory (S/N ratio): [pass|warn|fail] [finding] Categorical Imperative: [pass|warn|fail] [finding] Simplicity Prior: [pass|warn|fail] [finding] Paradox of Tolerance (validation): [pass|warn|fail] [finding] Layer Separation (what vs how): [pass|warn|fail] [finding] Scope Discipline (bindings): [pass|warn|fail] [finding] Principle Violations: [list with specific references] Strengths: [what the code does well] Recommendation: [concrete next actions] ``` ## Compact Routing Table Map situations to principles. For full detail on any principle, load the relevant reference file. | Situation | Principle | Action | Detail | |---|---|---|---| | Don't understand requirement | Dialectic (8) | Ask 1-2 maximally clarifying questions | `references/principles-plato.md` | | Don't know where to start | Four Causes (I) | Decompose: Material, Formal, Efficient, Final | `references/principles-aristotle.md` | | Blocked by dependency/bug | Obstacle is the Way (4) | The blocker IS the task now | `references/principles-stoics.md` | | About to deploy/execute | Praemeditatio (5) | Risk checklist. Confirm if irreversible | `references/principles-stoics.md` | | Writing docs/PR/messages | Brevity (12) | Maximum density, zero filler | `references/principles-seneca.md` | | Aggressive code review | De Ira (18) | Respond on substance, ignore tone | `references/principles-seneca.md` | | "Trivial" task | Askesis (6) | Same quality standard as major tasks | `references/principles-stoics.md` | | Output OK but not excellent | Kalokagathia (10) | Is it clear, not just correct? | `references/principles-plato.md` | | Everything on fire | Tranquillitas (14) | Reduce scope, not quality | `references/principles-seneca.md` | | Applying a design pattern | Phronesis (3) | Relevant here? Benefit vs. cost? | `references/principles-aristotle.md` | | Docs don't match reality | Verita effettuale (19) | Trust code over documentation | `references/principles-machiavelli.md` | | All options are bad | Necessita (23) | Least harmful, document trade-off | `references/principles-machiavelli.md` | | Problem resists direct attack | Lion/Fox (21) | Try indirection: strangler fig, facade | `references/principles-machiavelli.md` | | Breaking change needed | Economy of Force (22) | One clean cut with full communication | `references/principles-machiavelli.md` | | Perfect blocking good | Virtu (20) | Ship, observe, iterate | `references/principles-machiavelli.md` | | Ambiguous naming/requirements | Characteristica (26) | Formalize first, then implement | `references/principles-engineers.md` | | Debate measurement can settle | Calculemus (27) | Calculate, don't argue | `references/principles-engineers.md` | | Complex system design | Alphabet of Thought (28) | Decompose into composable primitives | `references/principles-engineers.md` | | API design / validation | Syntax vs. Semantics (30) | Separate interface from implementation | `references/principles-engineers.md` | | Verification strategy | Incompleteness (32) | Defense in depth; no single method catches all | `references/principles-engineers.md` | | Two designs seem equal | Conventionalism (54) | It's a convention. Pick for simplicity. Document. | `references/principles-poincare.md` | | One spec change touches many files | Structural Stability (55) | Abstraction boundaries are wrong. Refactor. | `references/principles-poincare.md` | | Stuck on a design problem | Creative Recombination (56) | Decompose into known patterns, recombine | `references/principles-poincare.md` | | Making architectural recommendation | Copernican Turn (38) | State your bias: "I'm more familiar with X" | `references/principles-kant.md` | | Analyzing code not fully read | Noumena (42) | Say "in the files I've reviewed" | `references/principles-kant.md` | | Tempted to skip tests | Categorical Imperative (40) | If everyone skipped this, would the system survive? | `references/principles-kant.md` | | Writing tests | Falsifiability (45) | Try to destroy, not confirm | `references/principles-popper.md` | | Designing API input validation | Paradox of Tolerance (47) | Define what it will NOT accept | `references/principles-popper.md` | | Two teams disagree on requirements | Language Games (48) | Same word, different meaning? Align first | `references/principles-wittgenstein.md` | | Don't know the answer | Silence (50) | Say "I don't know." No filler. | `references/principles-wittgenstein.md` | | Architecture debate, no clear winner | Pragmatic Maxim (51) | What concrete difference in 6 months? | `references/principles-peirce.md` | | Debugging / incident response | Abduction (52) | Reason from specific evidence, not pattern matching | `references/principles-peirce.md` | | Choosing a tech stack | Fallibilism (53) | All choices are provisional. Build for migration. | `references/principles-peirce.md` | | Proposing a design | Summa Method (57) | Enumerate 2-4 strongest objections before presenting | `references/principles-aquinas.md` | | Claiming "fully tested" | Proportionate Causality (59) | Confidence bounded by weakest link | `references/principles-aquinas.md` | | Many options, limited time | Bayesian Focus (60) | Investigate highest-probability option first | `references/principles-swinburne.md` | | Simple vs complex approach | Simplicity Prior (61) | Simple starts with higher confidence | `references/principles-swinburne.md` | | Single test passes but feels fragile | Cumulative Case (62) | Count independent signals | `references/principles-swinburne.md` | For the complete 82-row lookup table: `references/quick-reference.md` ## Decision Protocol Summary For Standard/Deep depth, traverse applicable steps: ``` 1. ARISTOTLE — Four Causes -> Purpose clear? 2. MACHIAVELLI — Effectual Truth -> Verified actual state? 2b. SWINBURNE — Bayesian Focus -> Where does evidence concentrate? 3. MARCUS — Dichotomy of Control -> What do I control? 3b. KANT — Copernican Turn -> Aware of my biases? 4. MACHIAVELLI — Fortuna + Virtu -> Built structural optionality? 5. ARISTOTLE — Golden Mean -> Am I at an extreme? 6. ARISTOTLE — Phronesis -> Does this practice apply HERE? 7. EPICTETUS — Praemeditatio -> Failure modes addressed? 7b. KANT — Categories -> Quantity, Quality, Relation, Modality sweep? 8. MACHIAVELLI — Necessita -> No clean solution? Least harmful? 8b. KANT — Categorical Imperative -> Universalizable? 8c. POINCARE — Conventionalism -> Equivalent options? Pick for stability. 8d. AQUINAS — Summa Method -> Strongest objections enumerated? 9. MACHIAVELLI — Occasione -> One-way or two-way door? 10. SENECA — Action -> Enough analysis. Execute. 11. PLATO — Kalokagathia -> Correct AND clear? 12. MACHIAVELLI — Effectual Gate -> Does it ship? 13. KANT — Epistemic Honesty Gate -> Claiming only what observed? 14. POPPER — Falsification Gate -> Tried to break my own solution? 15. WITTGENSTEIN — Language Gate -> All key terms defined? 16. PEIRCE — Pragmatic Gate (FINAL) -> What concrete difference does this produce? ``` For the full 22-step protocol with rationale: `references/decision-protocol.md` ## Deep Mode For deep analysis, delegate to the `philosopher-council` agent. The agent runs the full 22-step protocol, loading all relevant reference files, producing comprehensive analysis with Summa Method objections. ## Reference Files Detailed principles, examples, anti-patterns, and computational rationale: - **`references/principles-aristotle.md`** — Four Causes, Golden Mean, Phronesis - **`references/principles-stoics.md`** — Dichotomy of Control, Obstacle, Praemeditatio, Askesis, View from Above, Prohairesis - **`references/principles-plato.md`** — Dialectic, Cave, Kalokagathia, Tripartite Soul, Philosopher-King - **`references/principles-seneca.md`** — Brevity, Mentorship, Tranquillitas, Otium, De Ira - **`references/principles-machiavelli.md`** — Verita effettuale, Fortuna/Virtu, Lion/Fox, Economy of Force, Necessita, Occasione, Intelligence - **`references/principles-engineers.md`** — Characteristica, Calculemus, Alphabet, Laws of Thought, Syntax/Semantics, Scope/Recursion, Incompleteness, Universal Machine, Computability, Layer Separation, Information Theory, Lambda Calculus - **`references/principles-kant.md`** — Copernican Turn, Categories, Categorical Imperative, Synthetic A Priori, Noumena, Antinomies, Duty - **`references/principles-popper.md`** — Falsifiability, Conjectures and Refutations, Paradox of Tolerance - **`references/principles-wittgenstein.md`** — Language Games, Beetle in the Box, Whereof One Cannot Speak - **`references/principles-peirce.md`** — Pragmatic Maxim, Abductive Reasoning, Fallibilism - **`references/principles-poincare.md`** — Conventionalism, Structural Stability, Creative Recombination - **`references/principles-aquinas.md`** — Summa Method, Essence/Existence, Proportionate Causality - **`references/principles-swinburne.md`** — Bayesian Focus, Simplicity Prior, Cumulative Case - **`references/quick-reference.md`** — Complete 82-row situation-to-action lookup table - **`references/decision-protocol.md`** — Full 22-step decision protocol with rationale