🧹 Refresh docs and landing page

AGENTS.md

@@ -10,39 +10,16 @@ Check the [README](README.md) for how to get started. The [Makefile](Makefile) c
 
 Datadex is a minimalistic and functional open data platform to help communities get, transform and publish open data.
 
-### Data Pipeline Architecture
+### Principles
 
 - Simple and functional.
 - Low abstractions, no frameworks.
-- Each file is a self-contained dataset.
-- Rely on Makefile for orchestration.
+- UNIX philosophy.
+- One file, one dataset.
 - Datasets are stored in the `data/` directory.
 
-```python
-import polars as pl
-from pathlib import Path
-
-def A(dep: pl.DataFrame) -> pl.DataFrame:
-    return dep.with_columns(A = pl.col("x") * 10).select("A")
-
-if __name__ == "__main__":
-    dep = pl.read_parquet("data/source.parquet")
-    res = A(dep)
-    Path("data").mkdir(exist_ok=True)
-    res.write_parquet("data/A.parquet", complression="zstd", statistics=True)
-    print("✅ A.parquet written")
-```
-
-### Deployment Strategy
-
-- Datasets are uploaded to Hugging Face Datasets (`datonic` organization).
-- API is deployed as Hugging Face Space using Docker.
-- Web interface served statically from `web/` directory.
-- Everything versioned in git following "data as code" principles.
-
 ## Key Dependencies
 
-- **Polars**: Primary data processing library.
-- **httpx**: HTTP client for data fetching.
-- **pyarrow**: Parquet file handling.
-- **uv**: Python package and environment management.
+- `uv`: Python environment management and runner.
+- `polars`: Data processing library.
+- `httpx`: HTTP client for data fetching.

README.md

@@ -11,38 +11,34 @@
 
 <br>
 
-Datadex is a fully open-source, serverless, and local-first Data Platform to improve how [communities collaborate on Open Data](https://davidgasquez.com/community-level-open-data-infrastructure/). Why?
-
-- Increase your community's coordination and shared understanding.
-- Makes it easy to publish data products built by your community, for your community.
+Open-source, serverless, and local-first data platform for your community. Datadex helps [communities collaborate on Open Data](https://davidgasquez.com/community-level-open-data-infrastructure/), increasing the community's coordination and shared understanding by making it easy to build and publish data products, by your community, for your community.
 
 > [!NOTE]
 > The previous version of Datadex, which utilized Dagster and DuckDB, can be found at [this commit](https://github.com/datonic/datadex/tree/e0906b943bb35a4507fa24aa33494e9d7ceb6fef).
 
 ## 🚀 Implementations
 
-Check other [real-world production Open Data Portals](https://davidgasquez.com/modern-open-data-portals/) of the Datadex pattern in the following repositories:
+Datadex is a pattern, not only a project. Check [real-world production Open Data Portals](https://davidgasquez.com/modern-open-data-portals/) based on Datadex:
 
-- [LUNG-SARG](https://github.com/open-radiogenomics/lung-sarg). The Open Data Platform for Sustainable, Accessible Lung Radiogenomics.
-- [Datania](https://github.com/davidgasquez/datania/). An Open Data Platform at national level that unifies and harmonizes information from different sources.
+- [Datania](https://github.com/davidgasquez/datania/). Open Data Platform that unifies and harmonizes information relevant Spanish datasets from different sources.
+- [Filecoin Data Portal](https://github.com/davidgasquez/filecoin-data-portal/). The main open data portal around the Filecoin ecosystem.
+- [LUNG-SARG](https://github.com/open-radiogenomics/lung-sarg). Open Data Platform for Sustainable, Accessible Lung Radiogenomics.
 - [Gitcoin Grants Data Portal](https://github.com/davidgasquez/gitcoin-grants-data-portal). A Data hub for Gitcoin Grants data and related models.
-- [Filecoin Data Portal](https://github.com/davidgasquez/filecoin-data-portal/). A data portal for data related to the Filecoin network and ecosystem.
 
 ## 💡 Principles
 
-> [Make Open Data compatible with the Modern Data Ecosystem](https://handbook.davidgasquez.com/Open+Data).
+> [Make working with open data easy and accessible by using modern tooling and approaches](https://handbook.davidgasquez.com/Open+Data).
 
 - **Open**: Code, standards, infrastructure, and data, all public and open source. Rely on open source tools, standards, public infrastructure, and [accessible data formats](https://voltrondata.com/codex/a-new-frontier).
-- **Modular and Interoperable**: Easy to replace, extend or remove components of the pattern. Environment flexibility (your laptop, in a cluster, or from the browser) when running and when deploying (S3 + GH Pages, IPFS, Hugging Face).
-- **Permissionless**: Any improvement is one Pull Request away. Update pipelines, add datasets, or improve documentation. When consuming, there are no API limits, just plain files.
+- **Modular and Interoperable**: Easy to replace, extend or remove components of the pattern. Environment flexibility (laptop, cluster, browser) when running and when deploying (S3 + GH Pages, IPFS, Hugging Face).
+- **Permissionless**: Any improvement is one Pull Request away. Update pipelines, add datasets, or improve documentation. No API limits, just plain open files.
+- **Simple**: Static assets, batch jobs.
 - **Data as Code**: Reproducible datasets with declarative stateless transformations tracked in `git`. Data is versioned alongside the code.
-- **Glue**: Be a bridge between tools and approaches. E.g: Use software engineering good practices like types, tests, materialized views, and more.
+- **Glue**: Be a bridge between tools and approaches. Follow UNIX philosophy.
 
 ## ⚙️ Setup
 
-Datadex is a Python project. The easiest way to get started is using a Python virtual environment.
-
-If you hit any issue, please [open an issue](https:github.com/datonic/datadex/issues/new)!
+You can get started easily by setting up a Python virtual environment. If you hit any issue, please [open an issue](https://github.com/datonic/datadex/issues/new)!
 
 ### 🐍 Python Virtual Environment
 

pyproject.toml

@@ -7,8 +7,8 @@ readme = "README.md"
 license = { text = "MIT" }
 dependencies = [
     "httpx>=0.28.1",
-    "huggingface-hub[hf-xet]>=0.34.3",
-    "polars>=1.32.1",
+    "huggingface-hub[hf-xet]>=0.36.0",
+    "polars>=1.34.0",
     "pyarrow>=21.0.0",
 ]
 
@@ -27,4 +27,6 @@ module-name = "datadex"
 module-root = ""
 
 [dependency-groups]
-dev = ["ipykernel>=6.30.1"]
+dev = [
+    "ipykernel>=7.0.1",
+]