exmxc Workers

Cloudflare Worker implementation for the mcp.exmxc.ai executable intelligence node. The service is both:

1. an agent-discoverable REST/JSON intelligence API, and
2. a spec-compatible MCP JSON-RPC 2.0 server at POST /mcp.

The code keeps the ES-module Worker entrypoint (export default { fetch(request, env) }) and stores no secrets in source.

Agent Commerce Intelligence + AI Commerce Signal

The latest AI commerce release is served at GET /ai-commerce/signal and through ex.ai_commerce.signal.get; the immutable foundation remains available by release ID. It distinguishes shopping research, recommendations, selection, referrals, and delegated checkout. Experience sentiment must refer to using AI for the shopping task. Publication gates prevent thin anecdotal samples from becoming a sentiment reading.

The releases contain zero verified shopper episodes and zero instrumented agent transactions. Four sourced Adobe benchmarks remain survey and referral context, and the Google Trends series remains a separate normalized search-interest lane. The newest method pilot adds 11 reviewed signed-in app attempts from four buyer intents, including three retries, seven completed two-turn answers, six clean selections, one contradiction, and four failures. A ten-task provider-API check is reported on its own surface. GET /ai-commerce/selection exposes the app labels and private-archive digests, while raw personal chats stay private. This convenience sample does not support Agent Selection Share™ or a product quality ranking. The previous broad wallet packets and NKE/TJX instrument have been removed from the active tree; legacy API routes return HTTP 410 with a successor URL, and the old Webflow page URLs hand off to the new pages. No paid daily collection is scheduled. See method and operations.

AI Power Index — monthly ranked edition

The current ranked index is served at GET /ai-power/rankings and exposed to agents through ex.ai_power.rankings.get. It covers 50 companies with scores and ranks for 2026, 2027, and 2030. Query year=2030 to sort that forecast, and edition=2026-09-18 for the immutable inaugural edition. The ledger is /ai-power/rankings/editions; the fixed method is /ai-power/rankings/methodology.

GitHub stores dated editions and the locked Four Forces methodology. Webflow presents a complete static table enhanced with year sorting and an edition selector. Run npm run build:ai-power-rankings to generate the page code from the same implementation used by the API. See monthly operations. Earlier v1 exposure and v2 pilot material below remains separate historical compatibility output.

Canonical identity and build

• Entity: exmxc
• Domain: https://exmxc.ai
• Founder: Mike Ye
• Worker host: https://mcp.exmxc.ai
• Build version: 2.16.0
• Stable build date / last_updated: 2026-09-21

lib/registry.js is the single source of truth for entity metadata, build metadata, dataset registrations, callable data tools, content links, and federated registries.

Discovery endpoints

Canonical MCP discovery is served from https://mcp.exmxc.ai/.well-known/mcp.json; the apex https://exmxc.ai/.well-known/mcp.json is intentionally not served because the apex is Webflow on a DNS-only record. The transport endpoint is https://mcp.exmxc.ai/mcp.

• GET / — REST/MCP discovery document
• GET /.well-known/mcp.json — MCP discovery pointer
• GET /capabilities.json — generated capability inventory
• GET /.well-known/tool-registry.json — generated tool registry
• GET /.well-known/openapi.json — OpenAPI document with servers: [{ url: "https://mcp.exmxc.ai" }]
• GET /.well-known/manifest.json — generated manifest
• GET /.well-known/ai-plugin.json — plugin manifest pointing at the OpenAPI document
• GET /webmcp-power-lens.js — browser-native Power Lens tools for the top-level Webflow page
• GET /health — operational health check; uptime is not asserted in the payload and is observed through Cloudflare observability

All JSON responses use shared CORS headers. OPTIONS returns 204 with Access-Control-Allow-Methods: GET, POST, OPTIONS and Access-Control-Allow-Headers: content-type, mcp-protocol-version.

MCP JSON-RPC transport

POST /mcp accepts one JSON-RPC 2.0 request per HTTP request and returns an application/json response for synchronous tools. GET /mcp returns 405 with Allow: POST.

Supported methods:

• initialize
• notifications/initialized
• ping
• tools/list
• tools/call

Example initialize request:

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "initialize",
  "params": { "protocolVersion": "2025-06-18" }
}

Example tool call:

{
  "jsonrpc": "2.0",
  "id": 2,
  "method": "tools/call",
  "params": {
    "name": "ex.speg.get",
    "arguments": { "ticker": "NVDA" }
  }
}

tools/list, /capabilities.json, and /.well-known/tool-registry.json are generated from the same DATA_TOOLS registry so tool IDs and counts stay consistent.

Tool inventory

Callable JSON tools from DATA_TOOLS:

• ex.ai_commerce.signal.get
• ex.entities.get
• ex.speg.get
• ex.speg.index.get
• ex.datasets.index.get
• ex.ai_power_index.get
• ex.four_forces.get
• ex.ai_power.profiles.get
• ex.power_lens.v2.get
• ex.entity_in_a_box.get
• ex.power_lens.get
• ex.reality_gap.get
• ex.strategic_consequence.get
• ex.ai_power.analysis.top
• ex.eei.audit.run
• ex.convergence.latest
• ex.convergence.log

Content links from CONTENT_LINKS:

• ex.ai_commerce.page — https://www.exmxc.ai/ai-commerce
• ex.ai_commerce.methodology — https://www.exmxc.ai/ai-commerce-methodology
• ex.speg_index.page — https://www.exmxc.ai/speg-index
• ex.speg_index.methodology — https://www.exmxc.ai/speg-methodology
• ex.ai_power.page — https://www.exmxc.ai/ai-power-index
• ex.ai_power.methodology — https://www.exmxc.ai/ai-power-index-methodology
• ex.framework.get — https://exmxc.ai/frameworks
• ex.signal.get — https://exmxc.ai/signal-briefs
• ex.lexicon.get — https://exmxc.ai/lexicon
• ex.capital.get — https://exmxc.ai/capital
• ex.doctrine.get — https://exmxc.ai/leadership-doctrine
• ex.about.get — https://exmxc.ai/about-us
• ex.audit.page — https://www.exmxc.ai/audit
• ex.reality_gap.page — https://www.exmxc.ai/reality-gap
• ex.strategic_consequence.page — https://www.exmxc.ai/strategic-consequence

WebMCP Power Lens surface

The Webflow Power Lens page loads GET /webmcp-power-lens.js after exposing its existing search and render workflow through window.exmxcPowerLens. Compatible browsers discover two page-scoped tools:

• run_exmxc_power_lens — resolves one v2 pilot company or ticker and renders its evidence-backed profile or explicit unknown state.
• get_exmxc_power_lens_result — reads the Power Card currently visible on the page without changing state.

The browser layer does not create a second scoring model. The Worker remains the source of truth for the scoped mechanism, four anchored judgments or unknowns, evidence, confidence, and release metadata. Power Lens v2 produces no composite score or universal rank. The page tools do not use live market data, place trades, or provide investment advice.

Verify the browser contract with:

npm run verify:webmcp

REST data endpoints

Bundled datasets are imported directly into the Worker. Dataset updates require a Worker deploy.

• GET /ai-commerce/signal — latest or immutable release, optional release ID

• GET /ai-commerce/methodology

• GET /ai-commerce/releases

• GET /schemas/ai-commerce-episode-v1

• GET /schemas/ai-commerce-release-v1

• GET /consumer-intent/* — retired (HTTP 410, successor URL)

• GET /entities

  • Source: data/entities.json
  • Filters: industry, entity_type, posture, capability
  • Canonical entity-name field: company

• GET /speg

  • Source: data/speg_index.json
  • Filters: sector, scarcity_layer, ticker

• GET /speg/index/v1

  • Sources: data/speg_index_v1/methodology.json and the immutable RC1 authoring packet under data/speg_index_v1/releases/
  • Filters: query, decision, membership_state, release
  • Returns ten durable-scarcity profiles with C/S/E/P/D judgments, independent eligibility gates, source provenance, review state, rule sensitivity, and separately nullable valuation links
  • RC1 is a draft research constitution: five include recommendations, five watchlist decisions, zero released members, and no live membership mutation

• GET /speg/index/v1/profiles/{stable-slug}

  • Exact issuer profile lookup; unknown issuers return 404

• GET /speg/index/v1/methodology

  • Locked speg-index-scarcity-v1.0.0 construct, anchors, gates, confidence, evidence, limits, and cadence

• GET /speg/index/v1/releases

  • Release ledger with a null latest-published pointer while only RC1 exists as a draft

• GET /speg/index/v1/releases/{release-id}

  • Exact immutable release lookup; unknown releases return 404

• GET /schemas/speg-index-profile-v1

  • Source: schema/speg_index_profile_v1.schema.json

• GET /datasets

  • Generated dataset index

• GET /datasets/ai_power_index

  • Source: data/ai_power_index_dataset_v1.json; labeled legacy v1 compatibility output

• GET /datasets/ai_power_index/schema

  • Source: schema/ai_power_index.schema.json

• GET /datasets/four_forces

  • Source: data/four_forces_dataset_v1.json; labeled legacy v1 compatibility output

• GET /datasets/ai_power_methodology_v2

  • Versioned definition, Four Forces domains, ordinal anchors, evidence rules, confidence, and cadence

• GET /datasets/ai_power_profiles_v2?query=NVDA

  • 20-company pilot release with scoped profiles, explicit collection coverage, and no universal composite

• GET /schemas/ai-power-methodology-v2

• GET /schemas/ai-power-profile-v2

• GET /schemas/ai-power-source-manifest-v2

• GET /datasets/entity_in_a_box_v1

  • Source: data/entity_in_a_box_v1.json

• GET /datasets/entity_in_a_box

  • Compatibility alias for /datasets/entity_in_a_box_v1

• GET /datasets/reality_gap_index

  • Source: data/reality_gap_index_v1.json
  • Ten-company V1 evidence ledger comparing AI narrative and observed capability

• GET /datasets/reality_gap_index/schema

  • Source: schema/reality_gap_index.schema.json

• GET /reality-gap?query=AAPL&sort=gap_ascending

  • Filters: query, classification, sort, limit
  • Returns transparent component scores, classification, confidence, and dated official evidence

• GET /datasets/strategic_consequence_scenarios

  • Source: data/strategic_consequence_scenarios_v1.json
  • Six canonical counterfactuals with force deltas, scarcity adjustments, causal chains, bottlenecks, assumptions, and validation signals

• GET /strategic-consequence?scenario=power_binding_constraint&query=NVDA

  • Required filter: scenario; optional filters: query, limit
  • Returns experimental relative advantage and pressure across the 84-entity legacy v1 exposure universe; it does not use AI Power v2 profiles

• GET /schemas/strategic-consequence

  • Source: schema/strategic_consequence.schema.json

• GET /datasets/entity_registry

  • Stable identifiers and canonical classifications for 745 registered entities

• GET /datasets/entity_clarity

  • Longitudinal release ledger

• GET /datasets/entity_clarity/snapshots/latest

  • Latest dated ECI observations; 744 of 745 registered entities are observed on 2026-07-30

• GET /datasets/entity_clarity/changes/latest

  • Computed change ledger for the 744-entity matched panel

• GET /datasets/entity_clarity/releases/latest

  • Release metadata, QA disclosures, summary metrics, and industry aggregates

• GET /schemas/entity-registry

• GET /schemas/eci-observation

• GET /schemas/eci-release

• GET /schemas/entity-clarity-evidence-v2

  • Automated pilot contract separating delivery, declared access, five-dimension Entity Clarity, static-content adequacy, model-test status, and legacy diagnostics

• GET /analysis/ai_power/top?limit=10

  • Labeled legacy v1 weighted exposure ranking retained for compatibility

• GET /power-lens?query=NVDA

  • Labeled legacy v1 compatibility view

• GET /power-lens/v2?query=NVDA

  • Resolves a company in the 20-profile pilot and returns its mechanism, four anchored judgments or explicit unknowns, evidence, confidence, and release metadata
  • Never emits an AI Power composite, percentile, or universal rank

• GET /schemas/power-lens

  • Legacy v1 response schema

• GET /schemas/power-lens-v2

  • Source: schema/power_lens_v2.schema.json

• GET /audit/run?url=https://exmxc.ai

  • Collects website delivery and declared provider-purpose access evidence for a public HTTPS URL
  • Returns a deterministic five-dimension Entity Clarity v2.1 score when usable static HTML is delivered and labels the prior EEI score as a legacy website diagnostic
  • Leaves failed, restricted, or unsupported collection unscored; no human review is required
  • Reports a separate adequacy flag so thin static or JavaScript application shells can be identified without changing the score
  • Does not report independent model representation unless a separate recorded model test exists

• GET /schema

  • Source: schema/schema.json

• GET /definitions

  • Source: schema/definitions.json

• GET /index

  • Source: index.json, with total_entities computed dynamically from the bundled entities length

Dataset examples

Entity record example using canonical company:

{
  "entity_id": "ent_blackrock_...",
  "company": "BlackRock",
  "industry": "Financial",
  "entity_type": "Public Company",
  "posture": "Open",
  "capability": "High",
  "ecc": 86,
  "observed_at": "2026-07-30",
  "methodology_id": "eci-ecc-v1"
}

ECC means Entity Clarity & Capability. It is interpreted alongside posture and capability, not as a standalone performance ranking. A Blocked observation is assigned ECC 0 by definition. The website-audit EEI methodology is a separate measurement family and must not be substituted for ECI/ECC observations.

sPEG record example using the row-level date field:

{
  "entity_id": "nvidia",
  "company": "NVIDIA",
  "ticker": "NVDA",
  "sector": "AI Semiconductor",
  "scarcity_layer": "Compute",
  "date": "2026-07-16",
  "price_usd": 207.4,
  "forward_pe": 24.69,
  "forward_eps_growth_pct": 36.9,
  "scarcity_multiplier": 2.5,
  "speg": 0.27,
  "calculation_method": "forward_fiscal_eps_midpoint_proxy"
}

The active sPEG snapshot uses disclosed user-supplied July 16 closing prices and forward fiscal EPS ranges. It is a proxy dataset rather than licensed point-in-time NTM consensus data. The prior February snapshot is preserved at data/speg_index_2026-02-13.json.

The additive sPEG Index v1 does not overwrite that legacy ratio dataset. It measures durable economic scarcity as an ordinal, evidence-linked selection judgment across Constraint, Substitution resistance, Economic capture, Persistence, and Cost of defense. SDS is summed only when all five dimensions are known; it never changes PEG or an economic multiplier. RC1 contains no security pricing records, PEG values, economic-sPEG values, performance history, weights, buy/sell calls, or return claims.

Legacy AI Power v1 weighted exposure record shape (historical compatibility only):

{
  "entity_name": "NVIDIA",
  "compute_exposure": 10,
  "interface_exposure": 9,
  "alignment_exposure": 7,
  "energy_exposure": 8,
  "ai_power_index": 8.6
}

AI Power v2 does not convert this total. Its current staging release uses a 20-company entity–mechanism–market–time cohort, four anchored judgments or explicit unknowns, and no composite or universal rank. See docs/ai-power-v2-operations.md for the evidence workflow and release gates.

AI Reality Gap record example:

{
  "entity_name": "Apple",
  "ticker": "AAPL",
  "ai_narrative_score": 89,
  "ai_capability_score": 63,
  "reality_gap": -26,
  "classification": "narrative_outrunning_deployment",
  "confidence": "medium",
  "snapshot_date": "2026-07-21"
}

Reality Gap uses AI Capability Score - AI Narrative Score. Positive values mean observed capability leads narrative; negative values mean narrative leads the evidence currently visible. V1 scores only ten companies from dated official disclosures and never imputes a score outside that ledger.

Strategic Consequence Engine V1 is an experimental legacy model accepting one of six versioned counterfactuals. It remains pinned to the v1 Four Forces weighted exposure scaffold and does not use AI Power v2 judgments. Its relative 0–100 scenario exposure score is conditional on the selected scenario and is not a probability or expected return.

AI jobs signal endpoint

GET /api/ai-jobs-signal remains an experimental ADS endpoint backed by Anthropic for synthetic posting generation. Benchmark and signal responses include:

• data_provenance: "synthetic-llm-generated"
• disclaimer: "Postings are model-generated illustrations for ADS analysis, not scraped or verified labor-market data."

Required Worker secret

The ADS signal route requires an Anthropic Worker secret binding:

wrangler secret put ANTHROPIC_API_KEY

Do not place ANTHROPIC_API_KEY in wrangler.jsonc or source files.

The separate manual AI Power assessment workflow requires OPENAI_API_KEY as a GitHub Actions repository secret. Do not place that value in source files, workflow inputs, logs, or configuration files.

Wrangler configuration

The active deployed Worker config is:

• wrangler.jsonc — main mcp.exmxc.ai/* Worker with observability enabled

The legacy workers/root-discovery/wrangler.jsonc remains in the repo for reference only. Its apex exmxc.ai/.well-known/* route is intentionally abandoned because the apex is Webflow on a DNS-only record and canonical discovery is served by mcp.exmxc.ai.

Deployment

Deploys are automatic on every push to main through .github/workflows/deploy.yml; no manual wrangler deploy is required for normal releases. The workflow also supports workflow_dispatch for an explicit redeploy.

The deploy job deploys the canonical exmxc-workers Worker from wrangler.jsonc. The abandoned exmxc-root-discovery Worker is not deployed by CI because apex discovery is intentionally not served; canonical discovery lives on mcp.exmxc.ai.

After deployment completes, the verify live job waits for edge propagation and runs:

node scripts/live-acceptance.mjs

The live acceptance harness checks the production origins, including POST /mcp, canonical REST/MCP discovery on mcp.exmxc.ai, tool inventory consistency, dataset/schema endpoints, CORS, and health semantics. Apex discovery is probed for information only and never fails CI. The workflow fails if the canonical live result does not pass these checks.

One-time GitHub Actions secret setup is required in repo Settings → Secrets and variables → Actions:

• CLOUDFLARE_API_TOKEN — scoped Cloudflare API token for deploying Workers and routes
• CLOUDFLARE_ACCOUNT_ID — aeb064fbc195ef8f54ebce0f51897a63

Do not commit the Cloudflare API token or any other secret value to source.

Registry submission packet

The registry/ directory contains a generated MCP registry submission packet for external discovery directories. Regenerate it whenever lib/registry.js changes:

node scripts/build-registry-packet.mjs

The generator imports the canonical entity, build, transport, and DATA_TOOLS definitions from lib/registry.js, then rewrites:

• registry/packet.json
• registry/server.json
• registry/SUBMISSION.md

This keeps registry submission metadata aligned with the live MCP tool inventory.

Repository structure

worker.js                         Main Cloudflare Worker
lib/http.js                       Shared JSON/CORS response helpers
lib/registry.js                   Single source of truth for metadata, datasets, tools, links, federation
lib/queries.js                    Shared REST + MCP query implementations
lib/speg-index-v1.js              Deterministic sPEG Index score, gate, release, filter, and semantic validation engine
lib/ads-classifier.js             ADS classification helper
lib/ads-taxonomy.js               ADS taxonomy definitions
data/*.json                       Bundled datasets
data/speg_index_v1/               Versioned sPEG Index methodology and immutable authoring releases
schema/schema.json                Entity dataset schema with canonical company field
schema/definitions.json           Semantic definitions
schema/ai_power_index.schema.json AI Power Index JSON Schema
schema/power_lens.schema.json     Power Lens response JSON Schema
schema/reality_gap_index.schema.json Reality Gap dataset JSON Schema
schema/strategic_consequence.schema.json Strategic Consequence response JSON Schema
schema/speg_index_profile_v1.schema.json sPEG Index profile-release JSON Schema
index.json                        Static entity dataset index baseline
scripts/live-acceptance.mjs        Live deploy acceptance harness
scripts/build-registry-packet.mjs   Registry packet generator
registry/                           Generated MCP registry submission packet
webflow/power-lens-head.html       Staged Power Lens page head metadata and JSON-LD
webflow/power-lens-footer.html     Staged Power Lens responsive application bundle
webflow/reality-gap-head.html      Staged Reality Gap page metadata and JSON-LD
webflow/reality-gap-footer.html    Staged Reality Gap benchmark explorer
webflow/strategic-consequence-head.html Staged Strategic Consequence page metadata and JSON-LD
webflow/strategic-consequence-footer.html Staged Strategic Consequence responsive application bundle
webflow/speg-index-*.html          Staged sPEG Index hub, profile, and release templates; noindex until release approval
webflow/speg-methodology-*.html    Staged sPEG Index methodology page; noindex until release approval
workers/root-discovery/worker.js  Unused root .well-known MCP pointer Worker reference

MCP modernization notes (v2.7.0)

exmxc exposes a REST/JSON intelligence API plus an MCP server using Streamable HTTP on Cloudflare Workers. Tool and resource inventories are generated from lib/registry.js; public REST aliases and bundled dataset payloads are preserved.

Legacy Power Lens v1 + AI Reality Gap V1

GET /power-lens?query=... and ex.power_lens.get preserve the historical v1 weighted exposure view and are explicitly labeled legacy. The current evidence-profile surface is GET /power-lens/v2?query=... and ex.power_lens.v2.get; it returns scoped judgments or unknowns without a composite or universal rank. Reality Gap V1 remains a separate evidence ledger.

Strategic Consequence Engine V1

GET /strategic-consequence and ex.strategic_consequence.get share one deterministic implementation. V1 accepts only the six bundled canonical scenarios and performs no live retrieval, free-form scenario generation, hidden company overrides, or missing-evidence imputation. Every entity result discloses its Four Forces contributions and coverage for scarcity and Reality Gap modifiers. Leaderboards express relative conditional structural exposure inside the selected scenario; they are not investment recommendations or return forecasts.

ADS signal route

GET /api/ai-jobs-signal without signal parameters returns the deterministic public benchmark. Paid signal generation uses POST /api/ai-jobs-signal and requires Authorization: Bearer <ADS_SIGNAL_KEY>. The Cloudflare deployment should add a rate rule on this path as defense in depth.

Audit route

/audit/run and the ex.eei.audit.run tool validate targets before contacting the fixed upstream audit service. The paired audit build validates each HTTPS destination and redirect, uses a DNS lookup that rejects private, loopback, link-local, reserved, and mixed public/private answers, and limits redirects, response size, and request time. Deployment configuration should still apply conservative Cloudflare rate limiting to /audit/run and audit calls arriving via /mcp.

Structured content compatibility

Successful tool calls keep content[0].text as the JSON serialization of the complete handler result. structuredContent is only emitted for protocol-valid object results; top-level array results remain text-only to avoid introducing an unapproved wrapper.