fpf-problem-solving

First Principles Framework (FPF)

Attribution and license

This skill adapts First Principles Framework (FPF) by Anatoly Levenchuk, ailev/FPF, upstream commit b40c9df9e856c1c8c6a12f93ed8b310265007f68. The FPF specification text in sections/ is licensed under CC BY 4.0. Changes: the specification was split into section files and given generated navigation indexes; this SKILL.md adds an agent-oriented router and usage instructions. See the upstream licensing scope. Skill packaging and the splitter are MIT licensed. This adaptation is not endorsed by the FPF author.

An "Operating System for Thought" — a transdisciplinary architecture for reasoning, written in human- and machine-readable pseudo-code. FPF turns raw intelligence (human or machine) into organisationally usable reasoning: explicit bounded contexts, auditable artefacts, multi-view descriptions, and disciplined hand-offs between specialised actors.

Use cases

Use FPF whenever you need to think more rigorously than the situation's default.

Decompose a messy, cross-domain problem into parts that can be reasoned about independently
Make a high-stakes decision with incomplete evidence — and know what evidence is still missing
Get a mixed team to reason together without vocabulary collisions or hidden assumptions
Audit whether a conclusion is well-founded or just plausible
Transform or transfer an insight without losing its subject, precision, or stated limitations
Structure a proposal that must survive scrutiny from multiple expert perspectives
Generate alternatives systematically instead of anchoring on the first idea
Define what "better" means before comparing options
Classify what kind of problem you're facing before searching for solutions
Plan how an AI agent should select and sequence its tools under budget and trust constraints
Make a decision under uncertainty — identify options, weigh evidence, and commit with an auditable rationale
Reconsider a prior decision when new outcomes or changed conditions may invalidate its relied-on basis
Establish whether X causes Y — or just correlates — and determine what intervention would work
Publish a stable multi-view artifact without changing the source semantics
Assemble or check a public framework publication form while preserving product-specific bodies and references
Develop a holder system's capability for a named Work family and verify transfer in representative Work
Synthesize source ontologies for one bounded authoring decision without flattening their local meanings
Refresh a SoTA pack, benchmark, or evidence trail when evidence decays or telemetry changes
Synthesize architecture candidates, compare structural options, or repair modularity/reuse claims
Detect an ontic candidate, decide its first-use disposition, and govern whether a new concept or U-kind should be introduced instead of just renamed
Identify the exact system that acts or is intended to change, then trace a lost path from outside use to architecture, production, and recursive builders
Discover other systems that may bear relevant consequences before a decision closes
Explain how constituent actions perform larger work and whether a replacement preserves each receiving use
Qualify, trial, describe, transfer, compose, and introduce methods while checking their contribution and provision
Develop opportunities, attainable problems, and support configurations from current results
Fit checklist questions, criteria, and checking means to an occasion, then reconcile shared results
Recover a reusable method cautiously from several performances, logs, or observations without treating the evidence as the method itself
Make instructions usable after a change of speaker, recipient, location, time, or other reference condition
Choose the next action during ongoing work, resume interrupted work, or make applicable methods visible when current facts, authority, and stop conditions matter
Turn an ordinary handover, result demonstration, or successful workaround into a bounded question with a useful next move
Coordinate whether a system or episteme is being explored, shaped, evaluated, or operated without confusing that state with claim assurance
Preserve domain vocabulary while recovering which exact FPF value or relation a local term currently claims
Recover the actual performer/support configuration for Work and test what survives interruption, handoff, delay, or reconfiguration
Decide whether a proposed mandatory step, check, record, wait, or tool use changes a substantive result or only adds ceremony
Compare what a finite addition, replacement, or intervention contributes relative to the current configuration
Clarify what a claim about human, machine, organizational, or cultural “learning” actually says changed
Clarify what “development” or “evolution” means by naming the changed subject, continuity rule, posture, and value basis
Probe whether apparent capability loss is really an envelope, support, access, adaptation, enactment, or capability-change issue before choosing redevelopment
Revalidate affected decisions when a relied-on source changes and the full set of receivers is unknown
Request or reuse a bounded specialist result while preserving supplier authority and the receiving decision
Select and combine diagrams, models, records, or other representations for one exact use
Compare what structure a reader can correctly recover from an explanation under stated preparation, access, and time constraints
Turn build, buy, reuse, outsource, or AI labels into comparable complete ways to obtain one result
Check whether advice or an evidence demand helps the recipient’s decision and leads to a feasible next step
Apply one selected FPF pattern to a current question and stop at the first useful result
Construct a first model, apply an unfamiliar theory, recover an argument, or revise reasoning after a premise changes
Connect physical, mathematical, and computational contributions into an interpretable answer
Design a measurement relation and determine what its indications can actually resolve
Track object identity across observations, describe constrained configurations, and retain information needed to predict change
Develop new questions, reusable operations, and branching search from available results
Compare continuation rules using bounded recorded histories without exposing future outcomes
Test whether an evaluation can imitate improvement, rebuild affected evidence after evaluation changes, or improve the way further improvements are obtained
Develop a line of thought from working notes, preserving its grounds and the relations that make a later continuation useful

How to navigate

The use cases above help decide WHETHER to invoke FPF. The router below decides WHERE to go once invoked.

Step 1 — Match the thinking need to a starting point

What you need to do	Start here
Decompose and model a whole, track continuing objects, constrain configurations, or predict state changes	05 Part A → A.1 Holons, A.1.1 Bounded Contexts, A.14 Mereology; A.1.RI object reidentification; A.3.3.CC constrained configurations, A.3.3.TR state change, A.3.3.PI prediction information
Identify, trace, and discover the acting or changed system, find omitted Systems that may bear consequences, then locate the first unsupported dependency from outside use through recursive builders	05 Part A → A.1.SCR System Recognition, A.1.CSD Consequence-Bearer Discovery, A.1.STM System-Thinking Long Mantra; 03 FPF Readme → Recover a Lost Path
Assign system roles and responsibilities, recover what “role” means, check permission, or distinguish production work from the identity and completion of its product	05 Part A → A.2 System Role Kinds and Assignments, A.2.8.PER Permission; 11 Part E → E.10.ROLE Role Meaning Recovery; 07 A.V → A.15 Role-Method-Work Alignment, A.15.PROD Production Work
Recover, qualify, and develop methods from performances or documentary sources, trial intended contributions, check coherence and transfer, maintain variants, and introduce them into practice	05 Part A → A.3.1.MR, A.3.1.DR, A.3.1.QM, A.3.1.WT, A.3.1.CC, A.3.1.FT, A.3.1.MV, A.3.1.MI; A.3.2.DE method descriptions; 09 Part C → C.39.MF development focus, C.39.RQ situated requirements, C.39.RP repertoire, C.39.MC changed conditions
Recover what project, process, or case language directly refers to before modeling it	07 A.V → A.15.6 Project, Process, and Case Recovery
Steer, resume, and recovery-test Work by choosing the next action from current facts, resuming after interruption, making applicable methods noticeable, or recovering the actual performer/support configuration and probing interruption, handoff, delay, or reconfiguration	07 A.V → A.15.7 Situation-Responsive Work Steering, A.15.8 Work-Performance Configuration and Recovery Testing, A.15.8.MS methodological support, A.15.10 Resume Interrupted Work, A.15.11 Make Applicable Methods Noticeable
Screen mandatory work for operational relevance before requiring a step, check, record, wait, cue, or tool use	07 A.V → A.11.OP Decision-Relevant Least Action and Operational Parsimony
Request or reuse specialist results for one receiving decision while preserving the other practice’s authority	07 A.V → A.15.9 Bounded Result from Another Practice
Select and combine representations for one exact action or decision without treating them as interchangeable	09 Part C → C.37 Use-Bounded Representation Selection and Co-Use, C.37.MP method representation profiles, C.37.CF content/form comparison
Set boundaries on what statements mean, recover agent-relative references after a transfer, distinguish relations from their individuated occurrences, or derive a missing relation claim	06 Signature Stack → A.6.B boundary norms, A.6.P.RI agent-relative reference recovery, A.6.REL relation obtaining and occurrences, A.6.RCD relation-claim derivation, declarations, gates, duties, and evidence
Prevent category errors or reconcile ontology premises before extending the framework	07 Constitutional Principles → A.7 Strict Distinction, A.7.1 Consequence-Guided Ontological Problem Solving, A.7.2 Premise Reconciliation, A.7.CP Constructive Premise Compact
Evaluate confidence in a claim or artifact — including formality, scope, and reliability of the underlying knowledge	08 Part B → B.3 Trust & Assurance; 09 Part C → C.2 KD-CAL / F-G-R scoring, C.2.2 Reliability, C.2.3 Formality
Compose parts into wholes, recover how actions enact larger work, and check constituent-method replacements	08 Part B → B.1 Gamma algebra, B.1.5.EW encompassing Work, B.1.5.RS constituent replacement, B.1.5.CM composite Method qualification; 09 Part C → C.13 Compose-CAL, C.20 Discipline-CAL
Reason through or coordinate development — initiate inquiry from ordinary work, develop a line of thought through working notes, construct a model, use or compare theories, recover arguments or constructions, revise premises, bridge domain vocabulary, name whether a subject is explored/shaped/evidenced/operated, or develop a new question	08 Part B → B.5 Reasoning Cycle, B.5.PI inquiry from ongoing work, B.5.WN working notes, B.5.1 Explore → Shape → Evidence → Operate, B.5.2 Abductive Loop, B.5.3 Domain-Concept Bridge, B.5.4 concept recognition; B.5.FM first model, B.5.RA argument recovery, B.5.RC construction recovery, B.5.RR reasoning revision, B.5.TU theory use, B.5.TC theory comparison, B.5.QD new questions, B.5.SD substantive discussion
Enter and apply FPF: choose a starting point or practical entry, find results across a DPF suite, or apply one pattern to a first useful result	03 FPF Readme → Start Here, Practical Entries; 11 Part E → E.11.DSG Architectural Reference, E.11.PUA Pattern Use
Generate alternatives / construct comparable ways to obtain one result, explore solution space, and keep apparatus use bounded	09 Part C → C.38 Comparable Ways to Obtain One Result, C.39 result construction, C.39.RO reusable operations, C.40 branching search and §4.8.1 recorded-history replay, C.40.CD co-development of problems and solutions; C.40.OD open-ended development, C.40.OC subject characterization, C.40.OI opportunity inquiry, C.40.OCN bounded opportunities, C.40.PC worthwhile problems/contributions, C.40.DC support configurations, C.40.JF joint reachability, C.40.JD joint development, C.40.NU neighboring uses/means, C.40.NP attainable problems, C.40.OP stepping stones, C.40.RQ requalification; C.17 Creativity-CHR, C.18 Open-Ended Search, C.19 Explore-Exploit, C.19.2 Use-Bounded Apparatus Application
Measure and compare options, construct or repair measurement relations, and assess indication resolution	07 A.V → A.17-A.19 Characteristics, CSLC & SelectorMechanism; 09 Part C → C.16 MM-CHR, C.16.MR measurement relation, C.16.IR indication resolution, C.16.RM measurement repair; 13 Part G → G.9 Parity / Benchmark Harness
Resolve conflicts across stakeholders or values	10 Part D → Ethics, bias audit, conflict optimization
Unify vocabulary or synthesize source ontologies across teams or domains without flattening source-local claims	12 Part F → F.0.2 Semantic Synthesis, F.0.2.MA method contribution allocation, F.0.2.RC constructive coverage, concept sets, bridges, UTS, lexical continuity
Transform, document, publish, and reuse epistemes, views, or frameworks while preserving subjects and product-specific bodies	06 A.IV.A → A.6.2-A.6.4 episteme morphing/viewing/retargeting (separate exact arrow, bounded-use assertion, and current-case judgment), A.6.3.NAR narrative rendering, A.6.3.RT.OE operative expression; 11 Part E → E.4.CM Connected Methods, E.8.ML connected-method pattern languages, E.4.PFIP Publication Integration, E.11.PFP Publication Form Profile, E.17 Multi-View Publication Kit
Sharpen expression — repair vague wording, recover exact method/work relations and model/explanation uses, clarify what “learning,” “development,” “evolution,” “interest,” or “curiosity” means in the current claim, surface ambiguity, or restore precision of epistemic / measurement / architecture terms	06 A.IV.A → A.6.P.WMR Exact Relation Recovery, A.6.H Wholeness Unpacking; 11 Part E → E.10 model/explanation use, E.10.LRN, E.10.DEV, E.10.INT interest/curiosity, E.10.ARCH, E.17.EFP; 09 Part C → C.2.P, C.16.P, C.30.P
Decide, reconsider, appraise advice, or compare contributions under uncertainty — compare a finite configuration change to the current configuration, determine whether changed conditions alter a prior decision’s basis, structure options, weigh evidence, and commit with auditable rationale	09 Part C → C.11.CRC Configuration-Relative Contribution Comparison, C.11.MV method-option worth, C.11 Decsn-CAL, C.11.DBR Decision-Basis Reconsideration, C.11.DUA Decision-Useful Advice and Evidence Demands
Reason about time and change — distinguish state readings, trends, currentness, and intervention-sensitive change, or recover an actual temporal structure before testing coordination	09 Part C → C.27 Temporal Claim Adequacy, C.27.TA Temporal Aspect; 03 FPF Readme → ACTUAL-TEMPORAL-STRUCTURE
Establish causality — climb the causality ladder, construct or challenge a causal model, identify causal structure, and check realizability	09 Part C → C.28 CausalUse-CAL, C.28.CM causal-model construction/challenge, C.28.MR mechanism replacement
Connect mathematical and computational reasoning — assess model fit, transfer results, realize computations, repair physical connections, compare symmetry transformations, or construct boundary balances	09 Part C → C.29 Mathematical Lens Use, C.29.1 result transfer and symmetry, C.29.2 computational formulation, C.29.3 computational realization, C.29.BB boundary balance; 08 Part B → B.5.MPC physical/mathematical/computational connection, B.5.MPC.R repair
Describe architecture or structural views — characterize structure, unfold constraint-governed structure, produce adequate architectural descriptions and view types, triage cross-scope architectural residuals	07 A.V → A.22 STRUCT-CAL, A.22.CGUS; 09 Part C → C.30, C.30.AD, C.30.ASV, C.30.LCA, C.30.ILC, C.30.TFS-REL
Connect transformation flows without collapsing independent structures into one flow or project	11 Part E → E.18.NET Network of Transformation-Flow Structures
Synthesize architecture candidates or reconcile several non-isomorphic structures of one practice, then assess modularity/reuse or publish ADR-style projections	09 Part C → C.31 Modularity, C.32 Architecture Candidate Synthesis, C.32.MWA Practice Architecture, C.32.MWA.PP performance/provision arrangements, C.32.PAD, C.32.ADR, C.32.ADA
Assess structural information — compare what a reader can recover from an explanation under stated conditions, or check architecture capture, source return, equivalence, morphisms, or discovery adequacy	09 Part C → C.2.8 Extractable Structural Information, C.33, C.34, C.35
Model context-dependent or indeterminate states — represent superposed, probe-coupled, or viability-bounded behaviour	09 Part C → C.26 Quantum-Like Modeling Lens, C.26.1 Probe-Coupled Boundary, C.26.2 Enacted Distributed State, C.26.3 Viability-Envelope
Survey a discipline and build, ship, or refresh a reusable toolkit	13 Part G → G.1-G.13 SoTA kit, CG-Frame, dispatcher, benchmarks, shipping, telemetry refresh, dashboards, external interop; 09 Part C → C.21 Discipline-CHR
Classify a problem type, test whether a candidate is admissible for a kind judgment, or compare kind identity before claiming a cross-local correspondence	09 Part C → C.22 Problem-CHR, C.22.PFR Problematic-For Relation, C.3 Kind-CAL, C.3.2 Kind Judgment, C.3.3 KindBridge
Use checklists to select questions for a result, fit the occasion, qualify criteria and checking means, preserve questions in templates, act on answers, and reconcile shared checks	09 Part C → C.41, C.41.OF, C.41.QC, C.41.TF, C.41.UA, C.41.SR
Define quality attributes ("-ilities") as structured bundles	09 Part C → C.25 Q-Bundle; 07 A.V → A.17-A.19 Characteristics
Govern ontology — detect an ontic candidate, decide its first-use disposition, and determine whether a new concept or U-kind is warranted	11 Part E → E.24 Ontic Introduction Discipline, E.24.CD Ontic Candidate Detection and First-Use Disposition, E.24.UK U-kind Admission and Ontic Settlement
Probe or develop capability — distinguish apparent capability loss from envelope, support, access, adaptation, or enactment failures; when development is separately selected, test whether improvement transfers beyond an exercise	11 Part E → E.23 Quality Improvement Loop (no-change comparisons, changed-evaluation evidence, improving improvement methods), E.23.CAE Capability Access and Expression Differential Probe, E.23.CDI Developing Capability for a Named Work Family
Reason about cultural evolution — describe cultural-evolution engineering or repair cultural-evolution wording	09 Part C → C.36 Cultural Evolution, C.36.P Precision Restoration, C.36.RP sustaining and renewing shared work, C.36.MC method culture
Orchestrate agentic tool use under budgets and trust gates	09 Part C → C.24 Agent-Tools-CAL
Trace provenance and revalidate affected uses when a relied-on source changes, or detect refresh debt	07 A.V → A.10 Evidence Graph, A.10.1 Revalidate Affected Uses; 13 Part G → G.6 Provenance Ledger, G.11 Telemetry-Driven Refresh & Decay

For complex problems, follow paths across multiple sections — the router shows where to start, not where to stop.

Step 2 — Read the _index.md, then the sub-section

Open the _index.md of the target section folder — it lists all sub-sections with line counts and descriptions.
Read only the specific sub-section file you need.
Do NOT load entire sections. Pick the narrowest file that serves the user's question.

Step 3 — Apply in plain language

Use plain language for the user. Introduce FPF-internal names (U.Holon, Gamma, F-G-R) only when they add precision the user needs.

Step 4 — Compose findings across sections

When a problem draws from multiple sections:

State each pattern's contribution in one line (e.g., "Bounded Contexts gives us the parts; Trust Calculus scores our confidence in each").
If patterns from different sections appear to conflict, check for category errors via A.7 Strict Distinction — the conflict is usually a level confusion (role vs. function, method vs. work), not a real contradiction.
Synthesize in natural order: decomposition first (what are the parts?), then evaluation (how confident are we?), then resolution (what do we do about gaps?).
Do not just list FPF patterns — weave them into a coherent answer to the user's actual question.

Starter prompt (example — adapt to the user's actual role and need)

You have the FPF specification loaded. Help me structure my project / problem / programme. Use plain language for an engineer-manager. Propose: (1) bounded contexts / specialisations, (2) decision criteria, (3) key alternatives, (4) hand-offs, and (5) missing evidence or tests before commitment. Introduce internal FPF names only when they add precision.

Section INDEX

Structural reference. Each entry is a folder — read its _index.md first, then pick the sub-section. Counts follow upstream H2 headings; the Architectural Reference includes its content-free FPF.Reference:End boundary marker.

#	Section	Sub	When to use
01	Title page	0	Identify: title, authorship, version date, top-level identity.
02	Table of Contents	0	Navigate: locate a pattern, keyword, query cue, dependency, or neighboring section.
03	FPF Readme	12	Enter, onboard, and recover: choose a starting point or practical entry, understand what each part contributes, connect actions to encompassing work, connect transformation flows, recover a lost path from outside use to recursive builders, or locate licensing and reuse terms.
04	Architectural Reference	22	Orient: read philosophy, Architectural Rationale, shared source synthesis, profile choices, whole-combination conditions, uncertainty posture, and purpose/non-goals.
05	Part A — Kernel	38	Decompose, identify, discover, trace, assign, recover, and authorize: holons, bounded contexts, acting/changed-system recognition, consequence-bearing System discovery, outside-use dependency tracing, roles, permissions, candidate-Method recovery from Work evidence and documentary sources, qualification, trials, coherence, transfer, variants, introduction, method descriptions, transformers, method/work separation, object reidentification, constrained configurations, state-change rules, predictive information.
06	A.IV.A — Signatures	29	Set boundaries, recover references, derive relations, transform epistemes, and render: recover agent-relative expressions after transfer; distinguish relations from occurrences; recover exact method/work and under-specified service/access relations; derive needed relation claims; keep source, receiving episteme, arrow, use claim, work, and publication distinct; classify statements, construct operative expressions, or render structure faithfully.
07	A.V — Principles	50	Prevent confusion, remove ceremony, recover direct subjects, and steer, resume, or recovery-test Work: category errors, ontology premises, decision-relevant operational parsimony, project/process/case language, situation-responsive next-action choice, interrupted-work resumption, making applicable methods noticeable, performer/support configuration and recovery probes, production-work identity, completion criteria and separate closure authority, measuring, comparing, evidence graphs and changed-source revalidation, methodological support, bounded specialist results, mechanism suites, transformation-step constraint validity, independent-check gate decisions, constraint-governed unfolding.
08	Part B — Reasoning	43	Compose, evaluate, and coordinate development: structural views (STRUCT-CAL), aggregation (Gamma), composite Method qualification, substantive discussion, constituent actions and encompassing Work, use-preserving Method replacement, trust scores, emergence, inquiry from ordinary work, working-note development, development states, domain-concept bridging, reasoning cycles, first models, concept recognition, theory use/comparison, argument and construction recovery, reasoning revision, new questions, physical/mathematical/computational connections.
09	Part C — Extensions	124	Score, compare, search, and architect: epistemic quality, reader-extractable structural information, typed reasoning, measurement, configuration-relative contribution comparisons, reconsideration after a decision’s relied-on basis changes, comparable result routes, representation selection and co-use, decision-useful advice and evidence demands, decisions, bounded apparatus use, temporal/causal/math lenses, causal-model construction/challenge, architecture synthesis across non-isomorphic practice structures, structural adequacy, cultural evolution, measurement construction/repair, result transfer, computational formulation/realization, symmetry comparison and boundary balances, reusable operations, branching search, recorded-history replay without future-outcome leakage, and reproducible construct use; method worth, representation profiles, content/form changes, development focus and requirements, repertoire, opportunity inquiry and joint reachability, checklist construction and use.
10	Part D — Ethics	5	Resolve conflicts: ethical trade-offs, bias auditing, safety overrides, conflict optimization.
11	Part E — Constitution and Authoring	70	Enter, apply, clarify, probe, develop, govern, reuse, and publish: practical entry and pattern use, DPF-suite navigation, learning/development/evolution and interest/curiosity claim recovery, connected methods, pattern languages and framework publication forms, quality-improvement loops with no-change comparisons and affected-evidence rebuilding, capability access/expression probing and development for named Work, edition continuity, multi-view publication, transformation-flow networks, pattern quality, ontic/U-kind governance.
12	Part F — Unification	24	Synthesize and align: bounded semantic synthesis across source ontologies, method contribution allocation and constructive coverage, concept sets, sense cells, bridges with separate bounded-use claims and reliance basis, system-role descriptions, UTS, lexical continuity.
13	Part G — SoTA Kit	15	Harvest and refresh disciplines: SoTA Packs, CG-Frames, dispatchers, provenance ledgers, benchmark harnesses, shipping, telemetry refresh, dashboards, external interop.
14	Part H — Reserved	0	Reserve: preserve the upstream Part H position for future specification content.