Get help or support the project
The scorecard is free for every transit agency, and it will stay that way. Choose the path that matches what you need: use the public tools, take a scorecard to your board, or fund the shared infrastructure.
Free
Use the public tools
Find an agency, check a zip before publishing, or request a full one-off score.
Free
Take a scorecard to your board
Every agency page has a printable board one-pager, and the same report ships as a single self-contained file for a board packet or a grant application.
Free for every agency, for its own feed, with no account.
Paid
Board reports for a whole program
A state program, a technical-assistance center, a vendor, or a consultancy that prepares packets for many agencies can get every agency's board report in one archive, branded with its own name and logo, and refreshed monthly if it wants.
Prices, delivery, and the refund commitment are on the bundle page.
Paid
The scorecard history, as tables
Every dated check and every finding across the whole corpus as two Parquet tables with a data dictionary, provenance, and license, rebuilt monthly, for a vendor, a consultancy, a program, or a researcher who needs the corpus over time. The dated scorecards it is built from stay free.
Public infrastructure
Sponsor the open project
Help cover scheduled refreshes, hosting, and evidence reliability without affecting scores or placement.
What it costs to run
The core site is deliberately cheap. Static pages, a scheduled daily scoring run, and a read API together cost single-digit dollars a month today. A bounded operating envelope of roughly $75 to $200 a month would add capacity while preserving hard spending caps. Sponsorship at that level pays for:
- Hosting and bandwidth for the site, the read API, and the grade badges the project publishes.
- The scheduled daily data refresh: fetching and validating more than 2,600 configured feed records in the current worldwide coverage, still mostly in the US and Canada. The status page reports the exact current count and records when each run completed.
- Release monitoring and evidence reliability, including the private retention needed to reproduce a published check and to audit a remediation receipt if one is ever submitted.
- Optional hosted one-off scoring capacity, but only if repeated demand justifies a capped service with abuse protection. It is not enabled today; the local check and GitHub-backed request path remain available.
Sponsorship does not automatically start a cross-agency realtime archive or another expansion. New collection work needs a named user, a bounded plan, and a cost gate.
Every service above runs with a hard spending cap. If a cost line ever threatens to grow past this range, the design changes before the bill does.
What stays free, no matter what
Everything an agency touches: the scorecards, the plain-language fix guidance, the open data and API, the badges, local pre-publish check, request-backed one-off scoring, and feed-health alerts. Sponsorship adds capacity. It never moves an existing feature behind a paywall. Nothing is subtracted from the free tier to create anything else.
What sponsorship does not buy
- No influence over the scores. Grades, methodology, weights, and which agencies are listed are never for sale. The full rubric is public.
- No official status. This is an independent data-quality lens, not a compliance determination or a certification. It states what a feed publishes; it does not certify it.
- No paid placement or score influence. Coverage views may sort public results to help programs decide where to start, but sponsors cannot buy prominence, change a grade, or alter how an agency is framed.
Sponsors are disclosed on this page and in the repository, always. Funding that comes with a request to change any of the above is declined.
How to sponsor
Sponsor through GitHub, or get in touch first if you would rather talk it through or need an invoice.
Sponsor on GitHub Start a sponsorship conversation
Prefer email? Write to Chelsea Kelly-Reif.
What each tier is sized to
The GitHub tiers are sized to real cost lines, so a sponsor can see what a month buys. The money is one pot and the spending caps hold whichever tier it arrives through; nothing is earmarked.
- $5 a month
- The static core: the site, the daily scoring run, and the read API, which together cost single-digit dollars a month today.
- $25 a month
- The dynamic edges: the subscribe API, the badge and API worker, and the MCP endpoint. They run on per-request compute, about $15 a month at plausible volume.
- $50 one time
- About one month of on-demand scoring at demo-era volume ($20 to $60), the next capacity the project would add.
- $100 a month
- On-demand scoring of any GTFS URL from a web form, every month (ADR 0029). It is written and not switched on; this is its running cost.
No sponsors yet. The first ones will be listed here, with what their support pays for.