Continuous corpus and production learning

Every new component, repository, conversion and production outcome should improve the shelf, taxonomy, recipes and future recommendations.

Problem boundary

What this subsystem owns and what it must hand to the rest of Actionist.

What it owns

Component/repo refresh schedules
Observed adaptation cost
Production quality and incident signals
Asset/archetype reranking
Recipe and contract evolution

What research already found

AutoSaaS already describes a procedural learning loop
Current corpus ranking is mostly metadata rather than production evidence
The remembered 1.3M/850k/80k corpus still lacks a current authoritative path

Open first-principles questions

What production signals change ranking?
How often are component and repo sources refreshed?
How do we avoid taxonomy drift?
When is an asset deprecated or replaced?

Inputs

New source feeds
Conversion receipts
Build metrics
Client outcomes
Incidents

Outputs

Updated rankings
New recipes
Deprecations
Research priorities

Connected knowledge nodes

P03 · Curated capability shelf P04 · Autonomous repo-to-block foundry P05 · Living UI component layer P06 · Design taste and preference learner P12 · Deterministic composition planner P14 · Runtime, verification and release

Scientific research protocol

Every research owner uses the same evidence funnel. No padding is allowed when a denominator is inapplicable or evidence is unavailable.

Prior evidence auditRead the canonical spine, all domain-linked artifacts, the 17-industry catalogue and existing contradictions before searching again.

Commercial 100 → 10Define a denominator, map approximately 100 companies/products, and produce first-party dossiers for the strongest 10.

GitHub 100 → 10Search and dedupe approximately 100 relevant OSS projects, then deeply inspect the strongest 10 at the capability and architecture level.

Local estate joinInspect AutoSaaS, Great Library, SISOCRM, 21st stores and other domain-specific precedents already on the laptop.

Innovation 100 → 10Generate a wide divergent idea register, score it against the problem and retain the 10 strongest falsifiable innovations.

First-principles synthesisDecompose the irreducible problem, challenge inherited assumptions, reconcile contradictions and state the minimum sufficient design.

Contracts and experimentsReturn the proposed contract, dependency edges, unknowns, falsifiers, experiments, stop rules and decisions.

Independent verificationA separate verifier checks counts, links, source claims, overlap, contradictions and unsupported conclusions before promotion.

Evidence and agent outputs

Existing research remains source evidence. New agent packets are written to one owned workstream directory and linked here after verification.

Artifact	Location	State
AutoSaaS method	`/Users/shaansisodia/SISO_Workspace/SISO_Agency/apps/AutoSaaS/framework/autosaas-method.md`	existing
Decision ledger	`research/actionmodel-builder-research-2026-08-26/phase-3/outputs/decision-ledger.md`	existing
Master synthesis	`knowledge/00-MASTER-SYNTHESIS.md`	existing
Future verified agent packet	`research/workstreams/P15-continuous-corpus-and-production-learning/runs/<run-id>/`	planned