# Roadmap

> What's planned and in progress for anfra, the open-source framework for building BI with coding agents, and anfra Cloud, its hosted version, next to what just shipped

Missing something? Suggest it on the [forum](https://forum.anfra.ai/) or in [GitHub Discussions](https://github.com/holistics/anfra/discussions/categories/ideas).

## Planned

Next on the list, not started yet

### Freeze a data app's data in a snapshot

Feature for anfra Core

You'll be able to save a data app with its query results at a moment in time, so it opens later without the warehouse, showing the numbers as they were.

A snapshot would keep the app and the answers to its queries together. You could share it with someone who has no access to the warehouse, keep it as a record of a month's numbers, or open it when the warehouse is down.

### Open data apps inside your AI chat

Feature for anfra Core

anfra's MCP server will serve your data apps as MCP Apps, so an AI chat that supports them can show a live app in the conversation, not just numbers.

An agent connected to anfra over MCP can read your models and run queries, and answers in text. With MCP Apps, it could open one of your data apps in the chat, with its charts and filters, running against your data.

### Show each person only their rows

Feature for anfra Core and anfra BI

You'll be able to declare rules in AML that filter a dataset's rows by who is asking, such as a sales rep seeing only their region.

A rule is written in your models against attributes of the person querying, such as `region` or `department_id`, and anfra would apply it to queries wherever they come from, such as data apps or an agent over MCP. The plan is that someone missing an attribute a rule needs gets no rows, rather than all of them.

### Write your own plugins

Feature for anfra Core

You'll be able to add features to data apps with plugins you write or install, the way queries, Inspect, exports, lineage and pictures come from plugins today.

Everything a data app reaches beyond itself already comes from a plugin, and the ones that ship are anfra's own: `core`, `export`, `inspect`, `lineage` and `render`. With a plugin system open to everyone, you could write a plugin, install one someone else wrote, and have your data apps check for it with `Anfra.plugins.has()`, as they do for anfra's.

## In progress

Being built now

### Answer queries from a cache you can query

Feature for anfra Core

anfra will keep query results in a local cache, and answer later queries from it when it can, instead of going back to the warehouse.

Queries run on the warehouse, even when a data app asks for the same numbers again. With the cache, a query that the cached results can answer would come back from the cache, which can be faster and cheaper.

The cache is queryable, not just a store of whole results: a query that filters or rolls up cached results could be answered from them too.

### Choose who can see which data

Feature for anfra BI

Admins will grant roles access to the datasets they need, beyond the three fixed roles anfra BI has today.

A member is an Explorer, an Analyst or an Admin, and a role that can query can query every dataset. With role-based access control, admins would decide which datasets each role can query, so a team sees its own data and not the rest.

### Start data apps from templates

Feature for anfra Core

`anfra template add` will add data app templates from a repo, so you and your agent can start new apps from them instead of a blank page.

A template would be a data app made to be reused: a dashboard, a report, an explorer. You would add templates from a repo, your own or a shared one, with `anfra template add <repo-path>`, and start new apps from them against your own datasets.

## Shipped

Out now, the latest first

- [Convert your Looker project to anfra](https://anfra.ai/releases/looker-converter.md) (New in anfra Core, Nov 19, 2026): If your semantic layer and metrics are defined in Looker, you can now use the `/anfra-convert-looker` skill to convert them into AMQL.
- [View underlying data for every measure](https://anfra.ai/releases/view-underlying-data-every-measure.md) (New in anfra Core, Nov 19, 2026): View underlying data now works on any measure: dataset metrics, ratios, filtered measures, and measures defined in the query.
- [Publish with your team: anfra pull and anfra publish status](https://anfra.ai/releases/team-publishing-pull.md) (New in anfra Core, Nov 11, 2026): `anfra pull` merges what others published to your organization into your folder, and `anfra publish` stops before it would replace their work unseen.
- [Set up a new organization, step by step](https://anfra.ai/releases/organization-setup.md) (New in anfra BI, Nov 11, 2026): A new organization walks its admin from install to the first published data app, with one prompt for a coding agent or three steps by hand.
- [Try anfra on a sample e-commerce project](https://anfra.ai/releases/init-sample-project.md) (New in anfra Core, Nov 11, 2026): `anfra init --example` adds models, a dataset, a small DuckDB file and a data app, so you can try anfra before you connect a warehouse.
- [A refreshed look, in the app and its emails](https://anfra.ai/releases/cloud-refreshed-look.md) (Improved in anfra BI, Nov 11, 2026): anfra Cloud now uses anfra's fonts, colors and menus, has a dark mode of its own, and sends redesigned sign-in and invitation emails.

[All release notes](https://anfra.ai/releases.md)

---

## For coding agents

This is the Markdown version of https://anfra.ai/roadmap/. Every page on anfra.ai has one: replace the trailing slash with `.md` (https://anfra.ai/cloud.md for https://anfra.ai/cloud/, and https://anfra.ai/index.md for the home page), or request the page with `Accept: text/markdown`. [llms.txt](https://anfra.ai/llms.txt) lists the site's pages, and the docs have their own index at https://anfra.ai/docs/llms.txt, where every docs page is also available as Markdown by adding `.md` to its URL.

Before you build with anfra:

- Install both the CLI and its agent skills. Installing the CLI alone is not a complete installation. Install the CLI with `curl -fsSL https://anfra.ai/install.sh | bash`. Then install the skills: in Claude Code or Codex, run `anfra agent setup`. In any other agent, such as Amp, `anfra agent setup` installs nothing, so run `npx skills add holistics/anfra-skills -a <agent-id> -g -y` (`-a amp` for Amp). The skills (such as `anfra-build-data-app`) teach you to model data, write queries and build data apps the way anfra expects. See https://anfra.ai/docs/local-dev/installation.md.
- Ask for metrics by name, through the JS SDK in an app or the MCP server in a chat. Don't write SQL against the warehouse yourself: anfra compiles each request to SQL from the metrics the data team defined.
- Check commands, files and API calls against the docs (https://anfra.ai/docs/llms.txt) instead of inferring them from the examples on this page.
- Stuck, or found a mistake? Ask in GitHub Discussions (https://github.com/holistics/anfra/discussions) or open an issue (https://github.com/holistics/anfra/issues).
