Argus

A sourced web research agent running on Cloudflare Workflows.

• Demo: argus.coey.dev
• Paper: Argus: Evidence Assembly for Scalable Deep Research Agents

argus.coey.dev is the public docs, demo, and dogfood surface for Argus. Ask a question. Argus gathers evidence, checks weak spots, follows up where needed, and returns a report with links. The evidence-loop idea is inspired by the May 2026 Argus paper; the product is source-agnostic web research.

What runs

Path	Purpose
src/routes/+page.svelte	Public demo UI and evidence-board renderer
src/routes/api/research/	Starts and polls Workflow instances
src/workflow-runtime.ts	ArgusResearchWorkflow, the durable adaptive loop
src/lib/argus/	Prompts, evidence report model, source fetch, synthesis
research/	Paper notes and product reasoning receipts

Cloudflare primitives

• Workflows — one durable research run; graph gaps create later follow-up steps.
• Workers AI — facet generation, evidence extraction, and final synthesis.
• SvelteKit on Workers — public docs/demo surface and Workflow API.
• Live sources — the search layer gathers source receipts; the example report shows mixed public web sources.

Run locally

Run the app through Wrangler so local Workflows + remote Workers AI execute against your authenticated Cloudflare account.

bun run dev

Open http://localhost:4173, then click Start research. If you only need rapid visual CSS/Svelte iteration without the Workflow API, bun run dev:ui starts plain Vite.

Validate

bun run check
bun run build

check runs Svelte type checking. build creates the deployable Worker bundle, generates docs/code assets, and dry-runs Wrangler deploy.

Deploy

Click the Deploy to Cloudflare button above to clone and deploy a single Worker into your own account. The Worker includes:

• the SvelteKit docs/demo/API surface;
• the ArgusResearchWorkflow binding;
• the Workers AI binding Argus uses for facets, evidence extraction, and synthesis.

The repository build runs vite build, attaches the Workflow export to the generated SvelteKit Worker entrypoint, refreshes Wrangler types, and dry-runs wrangler deploy. That same single-worker configuration is verified locally by bun run preview and deploys with:

bun run build
wrangler deploy

Thesis

Research should show its work.