Network intelligence layer

The public archive of economic denial.

CopeCheck tracks the gap between structural reality and public narrative: AI displacement, wage-circuit failure, policy theatre, labour-market exposure, and the comfort stories elites tell while the substrate moves under them.

The question is no longer whether disruption exists. The question is who can still describe the mechanism without flinching.

01

All CopeCheck Sites, One Spine

Each subdomain stays useful on its own. Codex makes the network legible as one system.

02

RSS Intake Is the Radar

Google Alerts feeds watch the surface. Search fallbacks fill gaps. OpenClaw turns the incoming stream into structured jobs.

03

Doom Market Signals

Prediction markets, UK labour data, and curated public polling become a live radar layer for questions the thesis cares about.

Waiting Market snapshot
CopeCheck Doom Index Market pulse

Prediction-market snapshot pending.

ONS Reality Check

UK labour baseline

YouGov Watchlist

curated, attributed
04

Network Evidence Surface

Stable links to each evidence layer. Current scores and counts live on their source sites.

Main Figure Index

Live source

    UK Evidence

    Live source

      Policy Catalogue

      Curated snapshot

        Career Exposure

        Historical snapshot

          Layoff Evidence

          Non-additive ledger
            05

            Latest Stories

            Fresh RSS captures from the main AI displacement feed, UK monitor, and layoffs scanner. Open a source or send it straight into the scoring queue.

            06

            Published Oracle Summaries

            Clean, high-confidence scores become permanent Codex entries with source links, attribution, score, rationale, and the route they belong to.

            Auto-publish gate
            07

            The Oracle Protocol

            CopeCheck does not score whether a claim sounds optimistic or pessimistic. It scores whether the claim survives contact with the mechanism.

            08

            Factory-Light Backend

            OpenClaw can do the reading and first-pass scoring. The backend keeps the record clean.

            Cheap where possible. Careful where it matters.

            A configured scoring worker performs the main pass, and the selected model is recorded with each job. The site stores receipts, attribution strength, rationale, confidence, and review status for every result.

            • Direct attribution can auto-publish above threshold.
            • Automatic feed publication holds weak or anonymous attribution for review.
            • Manual review-marked submissions carry a visible warning if published.
            • Every score keeps the source, quote, mode, and rationale.
            • Channels become filters over one intelligence graph.
            09

            Submit a URL

            Paste an article. Codex detects the destination, sends it to the scorer, prints the verdict here, and publishes to the right CopeCheck site when the evidence gate passes.

            Custom GPT Ask the Oracle