First Principles Framework (FPF)

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
• Transfer an insight across domains without losing precision or introducing category errors
• 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
• 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
• Refresh a SoTA pack, benchmark, or evidence trail when evidence decays or telemetry changes

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

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

1. Open the _index.md of the target section folder — it lists all sub-sections with line counts and descriptions.
2. Read only the specific sub-section file you need.
3. 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:

1. State each pattern's contribution in one line (e.g., "Bounded Contexts gives us the parts; Trust Calculus scores our confidence in each").
2. 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.
3. Synthesize in natural order: decomposition first (what are the parts?), then evaluation (how confident are we?), then resolution (what do we do about gaps?).
4. 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.