https://github.com/speakeztech/crqcindex
A Fidelity Conclave project to track public info on the approach to CRQC
https://github.com/speakeztech/crqcindex
Last synced: 20 days ago
JSON representation
A Fidelity Conclave project to track public info on the approach to CRQC
- Host: GitHub
- URL: https://github.com/speakeztech/crqcindex
- Owner: speakeztech
- Created: 2026-04-12T22:12:45.000Z (2 months ago)
- Default Branch: main
- Last Pushed: 2026-04-12T22:53:46.000Z (2 months ago)
- Last Synced: 2026-04-13T00:34:56.825Z (2 months ago)
- Language: F#
- Size: 122 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# CRQC Index
Monitoring convergence toward cryptographically relevant quantum computation.
The CRQC Index is a signal detection and analysis system that tracks progress toward cryptographically relevant quantum computers (CRQCs) across multiple cryptographic primitives. It provides per-primitive Z estimates (time to CRQC) with confidence intervals, helping organizations understand their exposure window under Mosca's theorem.
## A Conclave Project
The CRQC Index is built on [Conclave](https://speakez.tech/blog/conclave-a-speakez-platform-service/), SpeakEZ's managed platform for agentic AI systems on Cloudflare's global edge. The system implements a multi-agent architecture where specialized actors handle distinct responsibilities:
- **Prospero**: Coordinator Durable Object that owns the canonical Z estimate
- **Sentinel agents**: Source-specific scrapers (arXiv, patents, government advisories, vendor announcements)
- **Analyst agent**: Signal scorer and classifier, computes Z deltas per primitive
- **Librarian agent**: Deduplication, citation graph maintenance, provenance tracking
These actors communicate via Conclave's tell-first messaging pattern, with capability-based security enforced at the platform level through Fidelity.CloudEdge bindings and Cloudflare's Worker Loader API.
## Project Structure
```
crqcindex/
├── site/ Partas.Solid frontend (F# + SolidJS via Fable)
├── workers/ Cloudflare Workers (F# via Fable) - Conclave agents
├── cli/ F# CLI for Cloudflare deployment via Fidelity.CloudEdge
├── scripts/ Shell wrappers for CLI commands
└── docs/ Design documents and methodology
```
## Tech Stack
- **Frontend**: F# with [Partas.Solid](https://github.com/parta-solid/Partas.Solid) (SolidJS bindings), Vite, Tailwind CSS
- **Backend**: Conclave actor agents on Cloudflare Workers + Durable Objects + Sandboxes
- **Runtime SDK bindings**: `Fidelity.CloudEdge.Worker.Context` (workers-types), `Fidelity.CloudEdge.Agents` (agents-sdk), `Fidelity.CloudEdge.DynamicWorkflows`. Sandboxes binding pending. All four surfaces are required to compose the Conclave actor model — there is no partial-application path. See [docs/crqc-index-design-scaffold.md §"SDK Binding Foundation"](docs/crqc-index-design-scaffold.md) for the per-surface mapping.
- **Binding generation**: Standardized on Xantham per [Fidelity.CloudEdge/docs/00 Decision 7](../Fidelity.CloudEdge/docs/00_architecture_decisions.md). CRQC Index is the validation target for that decision.
- **CLI**: .NET 10.0 F# with [Fidelity.CloudEdge.Management](https://www.nuget.org/packages/Fidelity.CloudEdge.Management)
- **Deployment**: Cloudflare Pages + Workers, no Wrangler
## Prerequisites
- .NET 10.0 SDK
- Node.js 18+
- Cloudflare account with API token
## Getting Started
```bash
# Restore .NET tools (Fable)
dotnet tool restore
# Install npm dependencies
cd site && npm install && cd ..
# Development server
cd site && npm run dev:watch
# Build for production
cd site && npm run build
```
## Deployment
```bash
# Set environment variables
export CLOUDFLARE_API_TOKEN="your-token"
export CLOUDFLARE_ACCOUNT_ID="your-account-id"
# Full migration
./scripts/migrate.sh -v
# Or individual steps
./scripts/provision.sh -v
./scripts/deploy-pages.sh -v
./scripts/status.sh
```
## Site Sections (Planned)
- **Splash/Landing**: Visual Z-estimate instrument with per-primitive confidence intervals
- **Bulletins**: Short-form signal feed (hardware milestones, error correction, cryptanalysis)
- **Analysis**: Long-form synthesis connecting patterns across signals
- **Methodology**: Scoring rubric, signal taxonomy, Z estimation model
- **FUD Files**: Analysis of excluded signals (marketing inflation, hype amplification)
## License
Proprietary - SpeakEZ Technologies