Foundations

Type: kb/types/tag-readme.md

Core theory that the rest of the KB builds on. These notes define the quality criteria, the design methodology, and the fundamental constraints that shape every other decision.

Notes

• agent memory needs discoverable, composable, trusted knowledge under bounded context — unifying theory: three properties form the minimal artifact-quality basis for remembered knowledge that serves contextual competence
• context-efficiency-is-the-central-design-concern-in-agent-systems — context is the scarce resource; nearly every architectural pattern is a response to volume or complexity pressure
• A borrowed pattern transfers only as far as source and target share a mechanism — shared mechanism licenses transfer without independent justification, only over the layer where it reaches; analogy must earn adoption by target-side evidence
• short-composable-notes-maximize-combinatorial-discovery — the library exists for co-loading; short atomic notes maximize the surface area for cross-cutting discovery
• a knowledge base holds theories, descriptions, and prescriptions with asymmetric linking — content registers explain why theory, description, and prescription need distinct quality goals and asymmetric maintenance rules
• a universal knowledge framework demotes content taxonomies to defaults — content taxonomies demote to guarded defaults; what stays fixed is stipulated (answerability) or enforced (contract declaration), not certified universal; contradicts the tripartition's exhaustiveness claim
• A framework rule with a boundary-preserving rival is not an inherited constraint — the stable complement to the demotion claim: a one-way test where a boundary-preserving rival demotes a rule to a design choice, while finding no rival leaves the rule only undemoted, not certified
• a-knowledge-base-should-support-fluid-resolution-switching — KB quality measured by how fluidly it supports moving between abstraction levels
• charting-the-knowledge-access-problem-beyond-rag — exploratory decomposition: separates substrate, pointers, navigation modes, synthesis, and maintenance so RAG-vs-filesystem debates stop collapsing unlike tasks
• mechanistic constraints make Popperian KB recommendations actionable — bridges conjecture-and-refutation with bounded-context mechanics

Self-improving systems

• self-improving systems — evidence-responsive operative change to a system's own organization; reflective versus non-reflective self-improvement is the central distinction, and warranted autonomy is what costs
• actionable methodology — the prior vocabulary: a methodology supplies an intervention-relevant mapping; actionability relates it to an operator, operations, target, and setting

Rationale and design method

• first-principles analysis maps a design space before selecting within it — maps consequential alternatives before closure and tests claimed necessities against rival decompositions
• human analogies can motivate functions without determining component boundaries — retains useful functional hypotheses from human cognition while reopening their allocation across engineered components
• design rationale management in Commonplace — descriptive companion: how workshops, proposals, ADRs, and contracts distribute constraints, alternatives, and decisions—and what continuity shipped contracts do not enforce
• Derivation and inheritance give starting warrant; discriminating evidence or proof earns scope — derivation and inheritance give conditional or transferred starting warrant, an underdetermined choice gives none, and only discriminating evidence or proof earns the scope it actually covers

• Alexander's patterns connect to knowledge system design at multiple levels — (speculative) pattern language as document types, generative processes as codification

• agent context is constrained by soft degradation not hard token limits — the binding constraint is the soft degradation curve, not the hard token limit; agents are in the same soft-bound family as human cognition and organizational learning

• soft-bound traditions as sources for context engineering strategies — catalog of twelve traditions with transfer assessment: what's already working, what's plausible, what's aspirational

Other tagged notes

• A methodology governs its own extension only as far as it settles the meta-decisions it raises - A methodology governs its own extension only as far as it settles the meta-decisions its recommendations raise — representational form, verification, and authority
• A proposal-selection improvement loop requires search, evaluation, and operative retention - A proposal-selection improvement loop — candidates generated, evaluated with possible non-adoption, and accepted changes made operative — requires search, reject-capable evaluation, and operative retention
• A proximate target is checked for achievement, not for warrant - Between an improvement objective and its oracles sits a target level — a property pursued because it is held to serve the objective — whose linking claim no check in the loop evaluates
• A repeatable operative path keeps a redesign class open to revision - Operationalizes repeatable operative revision for a named redesign class as a causal path through representation, evidence-bearing determination, admission, installation, dependence, and continuity
• A retrieval miss is a local reflective-path failure - A missed relevant artifact leaves its represented aspect inert for the affected task and discovery route, while other loading paths and reflective aspects can remain causally connected
• Access burden and transformation burden are independent query dimensions - Queries have two independent difficulty axes — finding inputs (access) and producing the answer (transformation) — conflating them misroutes symbolic transformations through semantic processing
• Accumulation counts dependence through the retained result, not through the evidence it caused - Cumulativity counts dependence through the retained result only; counting the evidence channel that result caused would make it coextensive with operativity
• An adversarial human-agent loop can reconstruct the writing-is-thinking filter - The writing-is-thinking filter is the loop's, not the pen's — an adversarial human-agent loop can reconstruct what naive delegation loses, but only while the human stays the judge
• An omitted improvement-loop function and a frozen one need different repairs - Five proposal-selection systems expose frozen functions, while a direct-update contrast shows why absence of a gate is not omission; HyperAgents supplies a preliminary partial unfreezing
• Behavior-determining organization - Definition — the retained structure that shapes how a system will operate: parameters, policies, memory, rules, workflows, tools, code, architecture, evaluators; outputs and environment excluded
• Compounding is tested in later improvement, not by the accepting metric - Compounding evidence must come from later improvement episodes through displaced productivity measures and causal traces, not from the metric that accepted the earlier change
• Computationally directed self-improvement is a fixed-boundary reallocation ending in contraction - The progress question for self-improving systems is not category membership but which decision-bearing functions humans still supply; the endpoint test is whether the boundary can be contracted to exclude them
• Evidence bearing on an improvement objective - Definition — evidence bears on an improvement objective when it carries information about the criterion: gradients, rewards, errors, viability signals, tests, judgments; no evaluator required
• False-positive generation is filtered; false-positive acceptance becomes operative - False-positive generation faces evaluation before retention, while false-positive acceptance becomes operative and can compound
• Formal symbolic systems assess explanatory-reach only through causal and proof obligations - Formal symbolic systems assess explanatory-reach only after claimed generality is translated into causal or proof obligations inside a warranted model
• Gödel machines are a proof-governed case of reflective self-modification - The Gödel machine realizes reflective self-modification with a proof-gated acceptance rule, gaining model-relative rigor at the cost of excluding useful changes it cannot prove
• Improvements can accumulate without compounding - Improvements accumulate when later improvement consumes or preserves retained results; compounding requires an earlier benefit to counterfactually improve a later episode, directly or through reinvested savings
• Increasing computational autonomy relocates human effort to the frontier instead of reducing it - In an open-ended system, increasing computational autonomy need not cut total human hours — attention moves to the frontier — so measure improvements per human judgment, not human time
• LLM-executed methodologies are metacircular interpreters, not compilers - Self-hosting LLM methodologies are closer to metacircular interpreters than compilers: agents re-interpret natural-language rules each session, while stable paths codify into validators and commands
• Measuring autonomy well enough to see it improve is an open problem - Autonomy is reported per function rather than scored as a percentage, but that profile does not yet support comparison across systems or time
• Methodological and computational closure track different changes - Methodological closure tracks whether a retained method settles consequential decisions; computational closure tracks whether those decisions require a human actor
• Moving the interpretation–enforcement boundary requires cross-form coverage - Moving responsibility between model-interpreted rules and formal enforcement crosses natural-language and symbolic forms, so governing the transfer requires coverage of both and their mapping
• Operative change - Definition — a change is operative when it affects the system's subsequent operation over the relevant, declared horizon through a behavioral-authority path; operativity does not require permanence
• Parametric reproduction alone cannot replace an authoritative record - Reproducing a record's content does not transfer its authority. Replacement requires a governed artifact with stable identity, integrity, contestability, and attribution; mutable records also require currentness and addressable revision.
• Reach-assessment - Definition — judging whether a commitment's claimed explanatory-reach is genuine across natural-language, symbolic, and distributed-parametric forms
• Real self-improving systems occupy combinations no single rung captures - Casebook of thirteen placements on selected pathway-profile fields — Homeostat to Commonplace — reflection, cumulativity, allocation, and evidential limit in combinations no rung expresses
• Reflection buys addressability - Self-improvement can accumulate without reflection — parametric learners do — but non-reflective retention gives only indirect handles; reflective retention makes the changed object addressable
• Reflection makes retained lessons second-order: a lesson can reject or rescope a prior commitment - Reflection lets a retained lesson target a prior commitment explicitly — rejecting, revising, or rescoping it — while non-reflective correction acts indirectly through the substrate
• Reflective coverage is graded across representational forms - Reflective coverage is stated per represented form and operation profile; control of an external dependency does not make that dependency part of the system's reflective coverage
• Reflective system - Definition — a system is reflective relative to selected aspects when an internal process uses a causally connected self-representation of them in its operation
• Revising an improvement objective is licensed from outside it or is not improvement - Objective change is improvement only against a level outside both objectives; proxy revision, re-indexing, and surfaced under-specification subtract most apparent cases
• Self-improvement is relative to a declared objective - The improvement objective is a declared parameter alongside boundary and horizon, carrying two separable conditions — indexed by the analyst, antecedent in the pathway — whose failures differ in kind
• Self-improving system - Definition — operative, evidence-responsive change to a system's own behavior-determining organization, read against a declared boundary, horizon, and improvement objective
• Six reported self-improvement paths expose bounded redesign surfaces within supplied methods - Comparative evidence separates operative redesign, revision of governing machinery, and contributions to later improvement from declared editability across six reported self-improvement paths including HyperAgents
• Task-fitted structure costs cross-task reuse - Current task fit does not warrant permanent structure: harden inherited constraints, and keep task-derived choices replaceable until discriminating use demonstrates transfer
• Technical constraints turn KB objective-function choice from philosophy into engineering - Three technical constraints and the codification lever make KB objective-function choice testable engineering, not philosophy; goals set the loss, contracts and profiles specialize it, and oracle strength differs per profile
• The bitter lesson selects production methods, not representational forms - The lesson's axis is production method — hand-crafted versus search-and-learning — not representational form. Learned localized forms are therefore a coherent scaling hypothesis, with cross-artifact credit assignment as the decisive open problem
• The self-improving-system definition classifies its boundary cases without ad hoc exceptions - Ten boundary cases run against the self-improving-system definition — each classifies by the stated criteria alone; the stress they apply falls on boundary declaration, not on the membership clauses
• Theory-mediated self-improvement needs both interpretation and retention from one substrate - Reflective self-improvement inherits the theory-mediated sample-efficiency conjecture only where one substrate both interprets a theory about the system's own operation and retains it addressably
• Under sub-agent decomposition, feasibility is the heaviest fork's net load - Shows why decomposition changes feasibility from total operation cost to the largest residual load left on any fork after work is shifted to siblings or the parent
• Warranted autonomy is bounded by oracle domain - Bare autonomy is free, but warranted evaluation autonomy extends only to the candidates an oracle can assess with the required confidence
• World models assess explanatory-reach through action-conditioned prediction - Learned world models can assess explanatory-reach when action-conditioned predictions are tested across the interventions or shifts a commitment claims