Explore all 500 rules through readable foundation citations, one policy-use prompt per rule and source-bounded Saint Lucia and small-state cases. Then build a policy pathway and challenge its weakest condition.
Every case now shows its evidence date and method routes.
Policy prompts are shorter and keep the exact book rule beside the review question. Cases now expose their policy area, evidence cutoff, update question and bounded connections to PITONS, IIR, RIPPLE-4, SOT and Delivery State.
Explore rules, practise them on sample claims and record when you apply one. Progress describes the work you have done; it does not certify mastery or truth.
Rival-Explanation BudgetName at least one ordinary rival, one adversarial rival, and one data-generation artifact; specify what would distinguish each.
Your progress is stored only in this browser on this device. SLPA does not receive it. Clearing browser data may remove it. Claims typed into the reasoning check are never included.
Browser storage is available.
Saint Lucia evidence console 01
Put a fact inside its policy boundary.
10 reviewed files connect real Saint Lucia evidence to the rules. Each file shows what the source says, what it can support, what it cannot settle and the next policy question worth asking.
01Official factWhat the dated source reports
02SLPA calculationKept separate from official status
03Evidence boundaryWhat the record cannot establish
04Policy testThe decision still to be reasoned through
10of 10 files
LC-01 · People
Which population belongs in a planning formula?
Provisional official statisticReviewed 27 Aug 2026
Official record172,948de jure resident populationCensus night · 12/13 Mar 2022
What the source says
The Central Statistical Office reports a provisional, undercount-adjusted resident population of 172,948: 171,834 people in private households and 1,114 in institutions. The comparable 2010 census figure was 166,526.
Supports
A census-night national scale benchmark and a starting denominator for service planning, with its provisional status kept visible.
Does not establish
Saint Lucia's current population, a mid-year estimate, an exact unweighted headcount, or why the population changed.
Policy question
Which population benchmark and uncertainty adjustment should govern district service-funding formulas?
Primary sourceCentral Statistical Office of Saint Lucia
2022 Population and Housing Census: Provisional Results, Release 1
Three lanes stay deliberately separate: source-backed Saint Lucia cases, comparative precedents with a local transfer test, and a hypothetical policy-review prompt for every rule. A prompt is never counted as evidence.
Cross-context bridgesSame reasoning problem, different setting.
6 bounded comparisons
Bridge 01
How do connected assets become a reliable operating service?
A shared reasoning problem does not establish that the foreign instrument transfers.
Saint Lucia · Cost of living & measurement
Inflation slowed; the price level stayed high
04 rules
Case type
Source-bounded local case
Policy area
Cost of living & measurement
Evidence cutoff
CSO monthly CPI series through 2025; source pages reviewed 30 Aug 2026.
Source records
2
What the sources support
Using the CSO's April 2026 consolidated revised series, the all-items CPI rose from 100.58 in December 2019 to 119.87 in December 2025, a 19.2% increase. The revised annual-average inflation rate was 0.7% in 2025.
Inference the record does not support
Because inflation was only 0.7% in 2025, prices had broadly returned to their 2019 level.
Evidence boundary
These figures distinguish a price level from the rate of change. They do not measure the burden for a particular household, identify the causes of price changes, or show that every item rose by 19.2%.
Defensible policy use
Inflation slowed, but the revised national price index remained 19.2% above December 2019. Affordability policy should separately track the price level, household-specific baskets, and real income.
Turn a policy idea into an inspectable pathway from choice to mechanism, people, outcome, risk and indicator. Challenge a key assumption to see which downstream claims need fresh reasoning.
Practice console 04
Run a five-rule reasoning check.
Choose the kind of claim. The engine surfaces five questions from the complete library. It organizes your review; it does not declare the claim true or false.
This text is not saved with your progress.0/500
Step 01 · FrameGeneral claim
Check meaning, claim load, evidence path and your own standard.
Add a claim, or use the questions as a blank review sheet.
Step 02 · Examine0 of 5 questions reviewed
01Rule 001
Semantic validity
Do not call an argument invalid merely because a premise is false; instead try to construct a true-premise/false-conclusion model.
02Rule 447
Claim Load Principle
Before evaluating evidence, score what the claim is asking the evidence to carry; do not use one threshold for every claim.
03Rule 448
Precision Ceiling Rule
Round estimates to a level justified by measurement error, sampling error, model uncertainty, and data vintage.
04Rule 459
Evidence-Path Audit
Trace one decisive number backward until reaching the original measurement and document every judgment in between.
05Rule 500
Recursive Self-Application Rule
Swap labels and sides; if the evaluation changes, identify the relevant asymmetry or correct the double standard.
0 of 5 reasoning questions reviewed.
Complete library 05
Explore all 500 stable rules.
Move through four book parts and 30 sections, search exact source content, filter by mapped applications or on-device progress, and compare up to three checks.
A matching rule identifies a useful proof obligation; it does not establish PITONS conformance.02IIRAnalytical framework
Institutional Intervention Readiness
IIR separates an arrangement's current functions, harms, dependencies and unknowns before preservation, redesign, replacement or removal is judged ready.
What current function or dependency must be understood before change?
A matching rule can sharpen sequencing; it cannot rescue a course that fails a critical readiness floor.05DELIVERYDelivery model
Delivery State
Delivery State keeps authorization, financing, procurement, construction, acceptance, operation, maintenance and measured service results as distinct gates.
Which operational gate comes next, who owns it, and what observation would prove it?
A matching rule clarifies the next delivery test; it does not prove progress, completion or public benefit.
Analytical and delivery loop
IIR baseline→RIPPLE-4 candidate map→IIR readiness gates→SOT staged choice→Delivery and learning
PITONS spans the loop as the living record and assurance standard.
About the experiment
A model of the full engine, built in public.
The release tests how people move from a sourced fact to a bounded claim, an explicit policy pathway and a counterfactual stress test.
Now · v1.2
Follow the reasoning interlock
All 500 rules, concise non-evidentiary prompts, resolved foundation references and 48 dated cases now connect to the next relevant SLPA method without claiming a verdict.
Next
Deepen the strongest case files
Add claim-level source roles, primary-rule rankings and live update gates to high-use Saint Lucia cases while keeping foreign examples inside explicit transfer boundaries.
Later
Show reasoning provenance
Trace which rule affected which claim, assumption, evidence step or policy choice while preserving the human decision record.