/fpf-problem-solving

name: fpf-problem-solving
description: "Use First Principles Framework (FPF) to decompose problems, architect systems, evaluate alternatives, define quality, qualify and develop methods, initiate inquiry from ordinary work, disambiguate agent-relative instructions, steer or resume work, surface applicable methods, coordinate development states, bridge domain terms, screen mandatory steps, compare configurations, probe capability, decide under uncertainty, reconsider decisions after their basis changes, establish causality, reason about time, synthesize ontologies, govern ontic admission, publish views, revalidate sources, select representations, use checklists, and appraise advice. Triggers: FPF, bounded contexts, SoTA packs, assurance calculus, decision theory, operational parsimony, causal/temporal reasoning, architecture, modularity, transformation flows, structural adequacy, cultural evolution, lexical discipline, Parts A-I. Not for simple task planning, general philosophy, or Agile unrelated to FPF."

First Principles Framework (FPF)

Attribution and license

This skill adapts *First Principles Framework (FPF)* by Anatoly Levenchuk, [ailev/FPF](https://github.com/ailev/FPF), upstream commit `23b1cae396652ef1c1d63160aed163a3041c3ea1`. The FPF specification text in `sections/` is licensed under [CC BY 4.0](https://creativecommons.org/licenses/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](https://github.com/ailev/FPF/blob/main/LICENSING.md). 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” actual