A drop-in, source-cited federal-legislation data layer for your internal tool — bill tracker, policy-flow charts, and a per-bill view that links straight to Congress.gov, wired to live data in one prompt. A working, deployed internal bill-tracking dashboard showing LIVE federal legislation: a tracker table of recent House and Senate bills and resolutions with status, sponsor, committee, policy area, and latest action — each row click-through to the genuine Congress.gov bill page — plus policy-area / chamber flow charts and a per-bill view with a cited narrative summary and structured claims. It pulls from ~766 U.S. Congress manifests in the last 30 days (House ~482, Senate ~284, verified live 2026-06-30), shows an honest partial-coverage notice when the result set is incomplete, and keeps itself current via a scheduled Beacon refresh. You bring Synorb credentials only when the project asks through CREDENTIALS FIRST.
A real, deployable Next.js (App Router) + Node project, not a mockup: (1) a server-side API route at /api/bills that reads SYNORB_API_KEY + SYNORB_SECRET from .env (server-only, never NEXT_PUBLIC_) and calls the Synorb REST API over the verified-live U.S. Congress / U.S. House / U.S. Senate streams; (2) a React BillTracker table component (sortable by status, sponsor, policy area, latest action, chamber) where every row links to its real congress.gov source_url; (3) a PolicyFlow chart panel (Recharts) showing bill volume by policy area and chamber over a chosen window; (4) a per-bill detail view rendering the Synorb brief + structured signal claims with the source clickthrough always present; (5) an honest empty/partial state that respects coverage.status; (6) a Vercel cron route backed by a saved Synorb Beacon that refreshes the dataset on a schedule. Runs with `npm run dev` locally and `vercel deploy` to a live URL.
It is the rare "drop-in data layer" that is actually drop-in: a govtech engineer clicks one prompt, answers a few plain questions, and ends up with a deployed, source-cited bill tracker on real Congress data — work that otherwise means standing up a Congress.gov scraper, normalizing bill status/sponsor/policy-area, and babysitting a refresh job. Synorb is the cited core (every row and claim resolves to a real congress.gov URL — nothing invented), Cursor's native stack does the rest (Recharts charts, a web-search MCP only for gaps clearly labeled as non-Synorb, optional MCP Apps cards), and the build-time MCP probe (mode="count", zero quota) means the UI is shaped by the true data, not a guess. Federal-not-state scoping is explicit, provenance is central, and it self-verifies by fetching and rendering live data with working links before it calls itself done.
No setup to learn. Cursor asks a handful of normal-person questions — each with a smart default, so “you pick” always works — then builds it with live Synorb data flowing in.
Smart default: Everything moving through Congress right now (you can filter by policy area in the app)
Smart default: Both the House and the Senate
Smart default: The last 30 days
Smart default: A clean, neutral govtech style — light theme, accessible, no party colors
Smart default: Yes — refresh once every morning
A neutral, accessible govtech reading room: a dense but legible bill-tracker table up top (status pills, sponsor, committee, policy-area tag, latest-action date), a compact PolicyFlow bar/area chart beside it, and a per-bill drawer that slides out with the cited brief, structured claims, and a prominent 'View on Congress.gov' source button. No party colors; status communicated by shape and label, not red/blue. Built for engineers who want to read the data, not a press release.
A Kanban-style bill-flow board where columns are legislative stages (Introduced -> Committee -> Floor -> Passed Chamber -> Public Law) and each card is a bill that moves left-to-right by latest action; a Recharts sankey/stacked chart shows volume flowing by policy area and chamber, and clicking any card opens the cited detail with the real congress.gov link. Great for compliance teams watching what's actually advancing vs. stalling.
An alert-first layout for monitoring: a saved-Beacon-backed 'what changed since yesterday' feed at the top (new bills, status changes in your chosen policy area), charts of weekly bill flow underneath, and the full searchable tracker below. Every change line and card carries its congress.gov source link, and a clearly separated 'web context (not Synorb-cited)' lane shows any web-search-MCP enrichment so provenance is never blurred.
INFRA-BACKED (dev codebase). A real Next.js/Node project deployed to Vercel. BUILD-TIME: Cursor installs the Synorb MCP via deeplink (https://mcp.synorb.com/mcp + token) so it learns the true data shape; it uses synorb-stream-search to lock the U.S. Congress / House / Senate stream_ids and synorb-manifests mode=count (zero quota) to size coverage, then a small compact=true pull to inspect real cited bill examples before writing UI. RUNTIME: a Next.js server/API route reads SYNORB_API_KEY + SYNORB_SECRET from .env (server-side ONLY, never NEXT_PUBLIC_) and calls the Synorb REST API; `npm run dev` locally, then `vercel deploy` to a live URL. A Vercel cron route + a saved Synorb Beacon refresh the dataset on schedule. Credentials are requested only at .env creation, with a link to https://synorb.com/connect; they never appear in client code.
Credential safetySynorb credentials stay server-side or local — never in client/frontend code. You only enter them when the agent asks, with credentials from synorb.com/connect.
Every tracker row, chart drill-in, and per-bill detail renders the manifest's real source.source_url (e.g. https://congress.gov/bill/119th-congress/house-resolution/1395) as a clickable 'View on Congress.gov' link — never a fabricated or guessed URL. The per-bill view shows the cited Synorb brief and structured signal claims directly above the source button. If coverage.status is partial, the UI displays an explicit 'showing N of M matching bills' notice and never claims to be complete; if a window returns zero bills, it shows an honest empty state with the filters that produced it. Any web-search-MCP enrichment is rendered in a visually separate 'web context (not Synorb-cited)' lane so Synorb-cited facts are never confused with web facts.
Yes. The build prompt connects Synorb at build time to learn the real data shape, then wires live Synorb data into the running product. A working, deployed internal bill-tracking dashboard showing LIVE federal legislation: a tracker table of recent House and Senate bills and resolutions with status, sponsor, committee, policy area, and latest action — each row click-through to the genuine Congress.gov bill page — plus policy-area / chamber flow charts and a per-bill view with a cited narrative summary and structured claims. It pulls from ~766 U.S. Congress manifests in the last 30 days (House ~482, Senate ~284, verified live 2026-06-30), shows an honest partial-coverage notice when the result set is incomplete, and keeps itself current via a scheduled Beacon refresh. You bring Synorb credentials only when the project asks through CREDENTIALS FIRST.
Only where they are safe. INFRA-BACKED (dev codebase). A real Next.js/Node project deployed to Vercel. BUILD-TIME: Cursor installs the Synorb MCP via deeplink (https://mcp.synorb.com/mcp + token) so it learns the true data shape; it uses synorb-stream-search to lock the U.S. Congress / House / Senate stream_ids and synorb-manifests mode=count (zero quota) to size coverage, then a small compact=true pull to inspect real cited bill examples before writing UI. RUNTIME: a Next.js server/API route reads SYNORB_API_KEY + SYNORB_SECRET from .env (server-side ONLY, never NEXT_PUBLIC_) and calls the Synorb REST API; `npm run dev` locally, then `vercel deploy` to a live URL. A Vercel cron route + a saved Synorb Beacon refresh the dataset on schedule. Credentials are requested only at .env creation, with a link to https://synorb.com/connect; they never appear in client code.
Every tracker row, chart drill-in, and per-bill detail renders the manifest's real source.source_url (e.g. https://congress.gov/bill/119th-congress/house-resolution/1395) as a clickable 'View on Congress.gov' link — never a fabricated or guessed URL. The per-bill view shows the cited Synorb brief and structured signal claims directly above the source button. If coverage.status is partial, the UI displays an explicit 'showing N of M matching bills' notice and never claims to be complete; if a window returns zero bills, it shows an honest empty state with the filters that produced it. Any web-search-MCP enrichment is rendered in a visually separate 'web context (not Synorb-cited)' lane so Synorb-cited facts are never confused with web facts.
A real app. Synorb is the cited data core; it also composes the platform's own tools: Next.js App Router + Node (Vercel infra-backed deploy), Synorb MCP installed via Cursor deeplink (https://mcp.synorb.com/mcp + token) at build time, Recharts (npm) for policy-area / chamber bill-flow charts, Vercel Cron Jobs (scheduled Beacon refresh route), Web-search MCP (gaps-only, e.g. enriching a sponsor's district — visibly labeled as non-Synorb), MCP Apps interactive in-chat bill cards (optional), .env server-side secrets (SYNORB_API_KEY, SYNORB_SECRET).
A few plain-English questions, each with a smart default so you can just say 'you pick': What kind of legislation should the tracker focus on — a policy area (like health care, defense, immigration, the environment) or everything moving through Congress?; Which chambers do you want to track?; How far back should the bill list and charts go?; How should the dashboard look?; Should it keep itself up to date automatically on a schedule?.
The same Synorb MCP connects to every major coding agent. Pick yours.
Generate credentials, connect Cursor, and ship something with live, cited intelligence.