Context7
Up-to-date docs for AI agents
Descripción del creador (original en inglés)
Search, filter, count and sum Australian government open data, with every version kept.
Estos tres nunca se combinan en una sola puntuación: las opiniones, las mediciones y la popularidad son cosas distintas.
Todavía no escribimos un resumen sencillo para este. Abajo están los datos que leímos del registro.
Elige la app de IA que usas y copia la configuración.
Necesitas
¿Qué app usas?
Elige tu app arriba. La recordamos, así que la próxima vez el botón dirá “Instalar en …”.
Funciona en Claude para web, escritorio y móvil:
https://publicdata.au/mcpEjecuta esto en tu terminal
claude mcp add --transport http publicdata https://publicdata.au/mcpPégalo en .cursor/mcp.json de tu proyecto, o en ~/.cursor/mcp.json
{
"mcpServers": {
"publicdata": {
"url": "https://publicdata.au/mcp"
}
}
}Tu navegador te pide permiso para abrir la app; luego la app muestra la configuración y te pide confirmar. Los valores de las claves quedan vacíos para que los completes tú.
Pégalo en .vscode/mcp.json de tu proyecto
{
"servers": {
"publicdata": {
"type": "http",
"url": "https://publicdata.au/mcp"
}
}
}Tu navegador te pide permiso para abrir la app; luego la app muestra la configuración y te pide confirmar. Los valores de las claves quedan vacíos para que los completes tú.
ChatGPT solo se conecta a servidores en internet:
https://publicdata.au/mcpLos nombres de los menús pueden variar un poco según el plan y con el tiempo.
Tu elección se guarda en una cookie en este dispositivo; no necesitas cuenta.
Esta configuración se genera a partir de la ficha del creador en el registro. Si el README dice otra cosa, sigue el README.
Todavía no hemos leído la lista de herramientas de este servidor; revisamos los endpoints por turnos.
Revisamos información que cualquiera puede ver; no hemos leído el código. Las revisiones que no aplican no se cuentan. · Revisado el 6 nov 2026
El nombre en el registro se verificó con el dominio publicdata.au (el registro comprueba quién es dueño del dominio)
Todavía no hemos leído este repo en GitHub
Todavía no hemos leído este repo en GitHub
Aún sin revisar; revisamos los endpoints por turnos
Aún sin revisar; revisamos los endpoints por turnos
Aún sin revisar; leemos la lista de herramientas cuando revisamos el endpoint
Comparado con todos los servidores activos de esta categoría en el registro. Las estrellas solo cuentan cuando quien lo publica es dueño del repo.
Superó 1 de 1 revisiones básicas aplicables. No hemos leído el código. Antes de agregar una clave o un token, revisa qué acceso pide el servidor y usa una clave que tenga solo los permisos que necesitas.
El registro no indica ninguna clave obligatoria, aunque algunos servidores te piden iniciar sesión la primera vez que los usas.
Se ejecuta en la nube del creador. Te conectas por internet sin instalar nada, pero lo que envías pasa por los servidores del creador.
El repo de GitHub cambió por última vez –, y la última versión (2.49.1) llegó al registro hace 14 días.
Las apps compatibles con MCP, como Claude Desktop, Claude Code, Cursor y VS Code. Se conecta por URL, así que también funcionan las apps que aceptan servidores MCP remotos, como ChatGPT en modo desarrollador.
En la misma categoría: Context7, Codebase Memory, Exa
Servidores de la misma categoría hechos por otros creadores, ordenados por estrellas.
Cómo leer una tarjeta:★ Estrellas de GitHubiTres pilares: reseñas · revisiones · usoiActualizado = fecha en que se publicó la última versión en el registroi
Up-to-date docs for AI agents
Codebase knowledge graph for AI agents — 162 languages, sub-ms queries, 99% fewer tokens.
Connect AI agents to Exa for web search, content fetching, and multi-step research.
Copiado tal cual del repo del creador. PopNx MCP no lo ha revisado ni lo respalda. Se quitaron HTML, imágenes y scripts. Los enlaces llevan fuera de este sitio.
Australian government open data, republished as CSV, Excel, JSON, Parquet, SQLite, DuckDB, GeoJSON and GeoPackage at URLs that never change. Every release the publisher makes becomes a dated version that keeps its content, with its schema, provenance, the publisher's own file and a diff against the release before.
This is an independent site run by National Digital. No government agency runs it, funds it or has endorsed it. Each dataset is republished under the publisher's licence with the attribution that licence asks for.
Contributions are welcome from anyone. A dataset from the backlog is one YAML file in register/, and a new file format is one writer in pipeline/publicdata/serialise/writers/. CONTRIBUTING.md has the steps for those and for adapters, fixes and the clients, and publicdata.au/contribute gives the overview. Every dataset page links to the register entry it is built from.
Taking part means agreeing to the Code of Conduct. SUPPORT.md says where to ask for a dataset or for help, and a vulnerability is reported privately as SECURITY.md describes. GOVERNANCE.md says who decides, and AGENTS.md is for coding agents.
register/ one YAML per dataset: source, licence with evidence, attribution, field allow-list, key, partitions, status. Backlog and blocked entries live here too.store/ one manifest per fetched version. Bytes are in R2 and pulled by hash before a build.pipeline/ the Python build: publicdata register validate | fetch | build | gate | split | store pull|push | dist-push | shards | purge.functions/ Cloudflare Pages Functions: votes, requests, saved dashboards, the query API, latest/ redirects, R2 fallback.explorer/ and package.json the explorer's browser libraries, copied into the build.clients/ the packages that call the site from other languages: clients/python is publicdata-au on PyPI and clients/r is publicdataau for CRAN. They take any slug, so a new dataset needs no release.docs/ARCHITECTURE.md the rules that decide the code.docs/CORRECTIONS.md how a fault in a published file is reported and fixed, and the log of every correction.docs/adr/ the decision records behind those rules: what each decided, what was weighed against it and what it costs.docs/ACCESSIBILITY.md the WCAG 2.2 AAA target, what CI enforces and the exceptions.CONTRIBUTING.md sets up a working copy, builds the fixtures, fetches a real dataset and walks through the routine changes. No credentials are needed for any of it.
.github/workflows/deploy.yml builds, gates, splits, pushes large files to R2 and deploys to the Pages project publicdata-au on every push to main, with a preview per pull request from this repository. Credentials sit in GitHub environments. production, which only main can use, holds the deploy's and the hubs'; preview holds a pull request preview's; fetch, also main only, holds the fetch's and the catalogue's. Each holds a Cloudflare token of its own, and these are the R2 buckets each token can reach:
| Environment | Workflows | Writes | Only reads |
|---|---|---|---|
production | deploy.yml on main | publicdata-dist, publicdata-build-cache | publicdata-raw |
preview | deploy.yml on a pull request | none | publicdata-raw, publicdata-dist, publicdata-build-cache |
fetch | fetch.yml, catalogue.yml | publicdata-raw | none |
The build cache, from which a deploy takes the versions it does not build again, has a bucket of its own, so that only the production deploy can change what the site publishes from it.
.github/workflows/fetch.yml reads every live source each Tuesday and the live feeds every other day, and opens one pull request per government whose sources changed. .github/workflows/catalogue.yml harvests the portals' dataset lists. Some portals turn away cloud addresses, so both run on a runner of our own in Australia. Its runner group admits only those two workflows on main, and they start only on their schedule or when a maintainer runs them. Both open their pull requests as the GitHub App "publicdata.au fetch", whose key is a secret only those environments hold. A ruleset lets only that app create or push data/ branches, so no person can open a pull request as the automation. .github/workflows/data-pr.yml approves those pull requests and nothing else.
The code is licensed under the GNU Affero General Public License v3.0 or later (LICENSE). If you run a modified copy as a service, you must offer its users the source of your changes. The clients in clients/ are under the MIT licence in each package's LICENSE.
The publicdata.au name and mark are not covered by the code licence; BRAND.md says what a copy must change.
The data is not ours to license. Each dataset is published under its publisher's licence, which its register entry, its manifest and every file name, with the attribution that licence asks for. The text we write for the register is under CC BY 4.0 (LICENSE-DATA.md). Test extracts, the brand typeface and the explorer's libraries are listed in THIRD-PARTY-NOTICES.md; the typeface is RandomMaerks' free package, redistributed under its own licence.