System Guide

DAEPOLIS is judgement infrastructure.

This guide describes the system as it is being built in code: decision records, claim ledgers, source hardening, review workflow, forecasts, exports, and calibration. It is intentionally shorter than the methodology page because the product should be inspectable before it is persuasive.

Core Surfaces

The public site should map to the database and workflow objects, not just explain the analysis style.

Decision Records

Durable records containing the question, decision owner, source pack, claims, evidence, forecasts, objections, review status, monitoring triggers, and later outcome.

Brief Forensics

Existing consultant reports, AI memos, ESG briefs, policy papers, board papers, and investment notes can be decomposed into claims, warrants, implied priors, belief debt, missing counter-evidence, falsifiers, adversarial objections, and questions before reliance.

Institutional Dependency Intelligence

Companies, countries, buyer markets, sources, claims, contracts, portfolios, regimes, public agencies, vendors, and decision records can be audited as risk-bearing dependencies with separate materiality channels, evidence gaps, monitoring triggers, rhetorical downgrades, and next actions.

Personal Intelligence Surface

The post-login Surface is becoming the live instrument: users should be able to add observations, source snippets, reviewer objections, and monitoring deltas to specific nodes while the system preserves provenance, freshness, review state, and the verification chain.

Canonical Claim Ledger

Atomic claims with support status, evidence links, verification state, disputes, retractions, and resolved forecast outcomes.

Source Hardening

Evidence states distinguish provisional, source-assisted, source-grounded, contradicted, and missing-source material.

Review Workflow

Candidate claims, evidence, and hardening tasks can be challenged, approved, rejected, or returned for source work.

Forecast Register

Predictions carry probability, resolution date, falsifier, public commitment payload, Brier score, and calibration bin.

Exports And API

Brief PDFs, production packages, audit-trail exports, and enterprise data resources expose the same record spine.

Build Status

DAEPOLIS is launch-stage infrastructure. Some surfaces are live; others are deliberately staged because audit systems become credible by refusing to fake maturity.

Live now

Workbench intake, source packs, source-hardening states, claim verification, public ledger, prediction register, PDF exports, review queue, cohort access framing, and public methodology pages.

Next

Surface Signal Intake, Brief-to-Surface memory import, Supabase freshness/migration-drift proof, editable candidate claims, approve/reject controls, framework-validation UI, audit-trail export, and delta-indicator definitions.

Runway

External timestamp anchoring, model-fit rejection surfacing, reference-class injection, probability distributions, red-team distributions, portfolio ingestion, and organization-scoped enterprise APIs.

Operating Principles
The record is the product; prose is only one view over the record.
Weak evidence should be marked, not hidden.
Forecasts are useful only when reality can resolve them.
Human review, disputes, corrections, and retractions are part of the system, not an embarrassment.
Public institutional names stay private until explicit Stage 2 permission is granted.
Design Partner Cohort

Autumn 2026

The cohort tests real institutional decision records with source packs, claims, evidence, review workflow, exports, and ledger accountability. For access, contact daepolisanalysis@outlook.com.

Apply for cohort