From ad7209a7f43071f7c57dd544a6f051bc1c5313d1 Mon Sep 17 00:00:00 2001 From: pazhik <133212006+MonteNegroX@users.noreply.github.com> Date: Thu, 13 Aug 2026 22:17:36 +0400 Subject: [PATCH 1/2] docs: add Russian README translation --- README.md | 5 ++ README.ru.md | 136 +++++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 141 insertions(+) create mode 100644 README.ru.md diff --git a/README.md b/README.md index d6e024d569..f606f1f74d 100644 --- a/README.md +++ b/README.md @@ -8,6 +8,11 @@ +

+ English · + Русский +

+ ### Open-source GenBI: generative BI for AI agents. *Your agents generate, deploy, and govern dashboards from any database, grounded in a context layer they can actually trust.* diff --git a/README.ru.md b/README.ru.md new file mode 100644 index 0000000000..6e84e4f59e --- /dev/null +++ b/README.ru.md @@ -0,0 +1,136 @@ +
+ + + + WrenAI + + + +

+ English · + Русский +

+ +### Open-source GenBI: generative BI для AI-агентов. + +*Ваши агенты генерируют, деплоят и управляют дашбордами из любой БД — на context layer, которому можно доверять.* + +**Wren AI — open-source generative BI (GenBI) engine: governed text-to-SQL и semantic-layer платформа на open AI context layer, 22+ источников данных.** + +[Docs](https://docs.getwren.ai) · [Discord](https://discord.gg/5DvshJqG8Z) · [Vision](https://www.getwren.ai/post/the-missing-context-layer-for-ai-agents-over-business-data) · [Blog](https://www.getwren.ai/blog) + +[![License: Apache 2.0](https://img.shields.io/badge/License-Apache_2.0-blue.svg)](https://github.com/Canner/WrenAI/blob/main/LICENSE) +[![PyPI](https://img.shields.io/pypi/v/wrenai?label=wrenai)](https://pypi.org/project/wrenai/) +[![GitHub Release](https://img.shields.io/github/v/release/Canner/WrenAI?logo=github&label=release)](https://github.com/Canner/WrenAI/releases) +[![Discord](https://img.shields.io/discord/1227143286951514152?logo=discord&label=Discord)](https://discord.gg/5DvshJqG8Z) +[![Last commit](https://img.shields.io/github/last-commit/Canner/WrenAI)](https://github.com/Canner/WrenAI/commits/main) +[![Follow on X](https://img.shields.io/badge/follow-@getwrenai-blue?logo=x&logoColor=white)](https://x.com/getwrenai) +[![Made by Canner](https://img.shields.io/badge/made_by-Canner-blue)](https://cannerdata.com) +![Stars](https://img.shields.io/github/stars/Canner/WrenAI?style=social) + +Canner/WrenAI | Trendshift + +
+ +> **2026-05-07**: Wren Engine вошёл в этот репо в [`core/`](./core). Старый `Canner/wren-engine` архивирован. Прежний GenBI app — ветка [`legacy/v1`](https://github.com/Canner/WrenAI/tree/legacy/v1) (**Wren GenBI Classic**). [Announcement →](https://github.com/Canner/WrenAI/discussions/2205) + +--- + +## Что такое WrenAI + +WrenAI — **open-source generative BI (GenBI) engine**: AI-агенты **generate, deploy, govern** BI — от governed **text-to-SQL** до shareable dashboard, 22+ data sources. + +Доверие даёт open **AI context layer** + governed **semantic layer (MDL)** — business semantics, definitions, examples, memory, governance + knowledge из docs/wikis/chat. + +![Wren AI generative BI architecture](./misc/wren-ai-architecture.png) + +## GenBI в трёх битах: Generate · Deploy · Know + +- **Generate.** Бизнес-вопрос → *governed* text-to-SQL и charts. +- **Deploy.** Dashboard на [`wren-core-wasm`](https://docs.getwren.ai/oss/sdk/wasm); ship на Vercel / Cloudflare Pages. +- **Know.** MDL, `instructions.md`, memory — reviewable, Git-friendly. + +## Почему agent builders выбирают WrenAI + +- GenBI end to end (answer → dashboard → URL) +- Knowledge as semantic layer (MDL), не «всё в prompt» +- Open by default (Apache-2.0) +- Correctness primitives: retrieval, dry-plan, structured errors +- Governed execution + context in Git +- На вашем warehouse / pipelines / semantic stack + +## Сравнение + +| | Raw LLM | Traditional BI | Bare semantic | **WrenAI** | +|---|:---:|:---:|:---:|:---:| +| Writes SQL | ✅ often wrong | ❌ | ❌ | ✅ governed | +| Business definitions | ❌ | partial | schema only | ✅ + non-schema | +| Dashboards | ❌ | ✅ manual | ❌ | ✅ agent-driven | +| Your agents | ✅ | ❌ | ❌ | ✅ | +| Open Git-friendly context | ❌ | ❌ | partial | ✅ | +| Governed 22+ sources | ❌ | per-connector | definitions only | ✅ | + +## Quickstart + +### 1. Install CLI + +```bash +pip install wrenai +pip install "wrenai[postgres,memory]" +``` + +Mainland China tip: `pip install wrenai -i https://pypi.tuna.tsinghua.edu.cn/simple` +HF: `export HF_ENDPOINT=https://hf-mirror.com` + +### 2. Discovery stub + +```bash +npx skills add Canner/WrenAI +``` + +### 3–6. Через агента + +1. Setup DB: *"Use Wren to set up my Postgres database."* +2. Enrich: *"Enrich my Wren project with the business context in `raw/`."* +3. Ask: *"Who are our top 10 customers by sales this quarter?"* +4. Deploy: *"Turn that into an interactive dashboard ... deploy it to Vercel."* + +Guides: [docs.getwren.ai](https://docs.getwren.ai) · sample `jaffle_shop`. + +```bash +wren skills get onboarding +wren skills get enrich-context +wren skills get genbi +wren query --sql '...' +wren ask "" --guided +wren ask "" --direct +``` + +## Semantic layer (MDL) + +Git-friendly definition of what data *means*: models, columns, relationships, views, cubes, metrics + `instructions.md` / `queries.yml`. + +## What's Included + +- MDL semantic layer +- Apache DataFusion engine, 22+ sources +- GenBI dashboards (`wren-core-wasm`) +- Knowledge & memory (LanceDB hybrid retrieval) +- SDKs: `wren-langchain`, `wren-pydantic` +- Governed execution primitives + +## FAQ (кратко) + +- **GenBI** — BI от AI-агентов: governed SQL + dashboards + context. +- **text-to-SQL** — да, governed; дальше dashboards. +- **semantic layer** — да (MDL) + AI context layer. +- **OSS vs Cloud** — OSS free (Apache-2.0); RLS/CLS, GenBI UI — commercial. [Open core](https://www.getwren.ai/en/open-core). +- **Sources** — BigQuery, Snowflake, PostgreSQL, ClickHouse, Redshift, Databricks, DuckDB, … [Connect](https://docs.getwren.ai/oss/guides/connect). + +## License + +Apache-2.0 — see [LICENSE](./LICENSE). + +--- + +

⬆ back to top

From f2d4e677d27adbe62d690de63a01a500ba836693 Mon Sep 17 00:00:00 2001 From: pazhik <133212006+MonteNegroX@users.noreply.github.com> Date: Fri, 14 Aug 2026 01:41:37 +0400 Subject: [PATCH 2/2] docs(i18n): complete Russian README localization --- README.ru.md | 192 +++++++++++++++++++++++++++++++++++---------------- 1 file changed, 133 insertions(+), 59 deletions(-) diff --git a/README.ru.md b/README.ru.md index 6e84e4f59e..8f6fdcdb17 100644 --- a/README.ru.md +++ b/README.ru.md @@ -11,13 +11,13 @@ Русский

-### Open-source GenBI: generative BI для AI-агентов. +### Open-source GenBI: generative BI для AI-агентов -*Ваши агенты генерируют, деплоят и управляют дашбордами из любой БД — на context layer, которому можно доверять.* +*Ваши агенты генерируют, деплоят и управляют дашбордами из любой базы данных — на context layer, которому можно доверять.* **Wren AI — open-source generative BI (GenBI) engine: governed text-to-SQL и semantic-layer платформа на open AI context layer, 22+ источников данных.** -[Docs](https://docs.getwren.ai) · [Discord](https://discord.gg/5DvshJqG8Z) · [Vision](https://www.getwren.ai/post/the-missing-context-layer-for-ai-agents-over-business-data) · [Blog](https://www.getwren.ai/blog) +[Документация](https://docs.getwren.ai) · [Discord](https://discord.gg/5DvshJqG8Z) · [Vision](https://www.getwren.ai/post/the-missing-context-layer-for-ai-agents-over-business-data) · [Blog](https://www.getwren.ai/blog) [![License: Apache 2.0](https://img.shields.io/badge/License-Apache_2.0-blue.svg)](https://github.com/Canner/WrenAI/blob/main/LICENSE) [![PyPI](https://img.shields.io/pypi/v/wrenai?label=wrenai)](https://pypi.org/project/wrenai/) @@ -32,105 +32,179 @@ -> **2026-05-07**: Wren Engine вошёл в этот репо в [`core/`](./core). Старый `Canner/wren-engine` архивирован. Прежний GenBI app — ветка [`legacy/v1`](https://github.com/Canner/WrenAI/tree/legacy/v1) (**Wren GenBI Classic**). [Announcement →](https://github.com/Canner/WrenAI/discussions/2205) +> **2026-05-07**: Wren Engine вошёл в этот репозиторий в [`core/`](./core). Старый `Canner/wren-engine` архивирован. Прежний GenBI app (Docker chat-first) — ветка [`legacy/v1`](https://github.com/Canner/WrenAI/tree/legacy/v1) (tag `v1-final`), теперь **Wren GenBI Classic**. [Announcement →](https://github.com/Canner/WrenAI/discussions/2205) --- ## Что такое WrenAI -WrenAI — **open-source generative BI (GenBI) engine**: AI-агенты **generate, deploy, govern** BI — от governed **text-to-SQL** до shareable dashboard, 22+ data sources. +WrenAI — **open-source generative BI (GenBI) engine**: AI-агенты **генерируют, деплоят и управляют** business intelligence — от governed **text-to-SQL** до shareable dashboard, на 22+ источниках данных. -Доверие даёт open **AI context layer** + governed **semantic layer (MDL)** — business semantics, definitions, examples, memory, governance + knowledge из docs/wikis/chat. +Доверие даёт слой снизу: open **AI context layer** плюс governed **semantic layer (MDL)** — business semantics, approved definitions, examples, memory, governance и company knowledge из docs, wikis и chat. Generative BI ровно настолько хорош, насколько хорош context; Wren — этот context, reviewable и reusable для любого агента. ![Wren AI generative BI architecture](./misc/wren-ai-architecture.png) -## GenBI в трёх битах: Generate · Deploy · Know +## GenBI в трёх тактах: Generate · Deploy · Know -- **Generate.** Бизнес-вопрос → *governed* text-to-SQL и charts. -- **Deploy.** Dashboard на [`wren-core-wasm`](https://docs.getwren.ai/oss/sdk/wasm); ship на Vercel / Cloudflare Pages. -- **Know.** MDL, `instructions.md`, memory — reviewable, Git-friendly. +- **Generate.** Агент превращает бизнес-вопрос в *governed* text-to-SQL и charts. Schema-aware retrieval, MDL planning, dry-plan validation и structured errors держат ответ корректным, а не «уверенно неверным». +- **Deploy.** Любой ответ → shareable browser-side dashboard на [`wren-core-wasm`](https://docs.getwren.ai/oss/sdk/wasm); ship на ваш Vercel или Cloudflare Pages одной командой. +- **Know.** Знание живёт в versionable, evidence-linked files: semantic models (MDL), company definitions (`instructions.md`), memory. Reviewable. Git-friendly. Не залочено в чужом UI. ## Почему agent builders выбирают WrenAI -- GenBI end to end (answer → dashboard → URL) -- Knowledge as semantic layer (MDL), не «всё в prompt» -- Open by default (Apache-2.0) -- Correctness primitives: retrieval, dry-plan, structured errors -- Governed execution + context in Git -- На вашем warehouse / pipelines / semantic stack +- **Generative BI end to end.** Governed text-to-SQL и дальше: answer → dashboard → URL, через агентов, которые у вас уже есть. +- **Knowledge management встроено.** Business meaning и approved definitions — reviewable **semantic layer (MDL)**, а не «всё в prompt». +- **Open by default.** Core, SDK и skills — Apache-2.0. +- **Correctness as primitives.** Rich schema retrieval, dry-plan, structured errors, value profiling, eval runner. +- **Governed execution.** Dry-plan, row limits, structured errors; definitions и examples в Git — reviewable, versioned, diff-able. +- **На вашем stack.** Warehouse, pipelines, existing semantic layer — не ещё один tool «вместо всего». -## Сравнение +## Как Wren сравнивается -| | Raw LLM | Traditional BI | Bare semantic | **WrenAI** | +| | «Сырой» LLM-агент | Классический BI | «Голый» semantic layer | **WrenAI** | |---|:---:|:---:|:---:|:---:| -| Writes SQL | ✅ often wrong | ❌ | ❌ | ✅ governed | -| Business definitions | ❌ | partial | schema only | ✅ + non-schema | -| Dashboards | ❌ | ✅ manual | ❌ | ✅ agent-driven | -| Your agents | ✅ | ❌ | ❌ | ✅ | -| Open Git-friendly context | ❌ | ❌ | partial | ✅ | -| Governed 22+ sources | ❌ | per-connector | definitions only | ✅ | +| Пишет SQL за вас | ✅ (часто неверно) | ❌ | ❌ | ✅ governed | +| Знает business definitions | ❌ | частично, в tool | ✅ (только schema) | ✅ + non-schema knowledge | +| Генерирует и деплоит dashboards | ❌ | ✅ (вручную) | ❌ | ✅ agent-driven | +| Работает через *ваших* агентов (Claude Code, Cursor, MCP…) | ✅ | ❌ | ❌ | ✅ | +| Open, reviewable, Git-friendly context | ❌ | ❌ | partial | ✅ | +| Governed execution на 22+ sources | ❌ | per-connector | ✅ (только definitions) | ✅ | -## Quickstart +## Wren для вас, если… -### 1. Install CLI +- Нужны **trustworthy BI**: answers *и* dashboards, не просто «правдоподобный SQL». +- Business logic (definitions, enums, units, joins) живёт **вне БД**, и агенты ошибаются. +- Нужен **AI context layer** и **semantic layer**, которые **open, reviewable, version-controlled**. + +**Пропустите Wren**, если нужен one-off chart из одного CSV или вас устраивает SQL без governance. + +## Быстрый старт + +WrenAI **agent-driven by design**: CLI + discovery stub, дальше ведёт AI-агент. Workflow guides внутри CLI, on demand. + +### 1. Установите CLI ```bash -pip install wrenai -pip install "wrenai[postgres,memory]" +pip install wrenai # core (DuckDB included) +pip install "wrenai[postgres,memory]" # add per-datasource and memory extras as needed ``` -Mainland China tip: `pip install wrenai -i https://pypi.tuna.tsinghua.edu.cn/simple` -HF: `export HF_ENDPOINT=https://hf-mirror.com` +> **Совет для пользователей в mainland China:** если `pip install` медленный или падает, используйте зеркало Tsinghua: +> ```bash +> pip install wrenai -i https://pypi.tuna.tsinghua.edu.cn/simple +> ``` +> Если HuggingFace model downloads time out: `export HF_ENDPOINT=https://hf-mirror.com` перед CLI. -### 2. Discovery stub +### 2. Discovery stub для AI-клиента ```bash -npx skills add Canner/WrenAI +npx skills add Canner/WrenAI # auto-detects Claude Code, Cursor, Cline, Codex, … ``` -### 3–6. Через агента +Stub ~50 строк: учит агента `wren skills get ` и `wren ask "" --guided|--direct`. + +### 3. Попросите агента настроить + +> "Use Wren to set up my Postgres database." + +Агент: `wren skills get onboarding` → connection profile, project scaffold, first query. + +### 4. (Опционально) Enrich — такт *Know* + +> "Enrich my Wren project with the business context in `raw/`." -1. Setup DB: *"Use Wren to set up my Postgres database."* -2. Enrich: *"Enrich my Wren project with the business context in `raw/`."* -3. Ask: *"Who are our top 10 customers by sales this quarter?"* -4. Deploy: *"Turn that into an interactive dashboard ... deploy it to Vercel."* +`wren skills get enrich-context` — grill или auto-pilot; пишет MDL, instructions, queries, memory. -Guides: [docs.getwren.ai](https://docs.getwren.ai) · sample `jaffle_shop`. +### 5. Вопросы — такт *Generate* + +> "Who are our top 10 customers by sales this quarter?" + +### 6. Dashboard — такт *Deploy* + +> "Turn that into an interactive dashboard I can filter and share, and deploy it to Vercel." + +[Build & deploy a GenBI app](https://docs.getwren.ai/oss/guides/genbi). Без своей БД: sample `jaffle_shop`. + +## Два такта сначала, потом третий ```bash -wren skills get onboarding -wren skills get enrich-context -wren skills get genbi -wren query --sql '...' +# Day 1 (agent-driven) +wren skills get onboarding # setup + first query (Generate) +wren skills get enrich-context # business context (Know) +wren skills get genbi # build & deploy dashboard (Deploy) + +# Day-to-day +wren query --sql '...' # query through MDL wren ask "" --guided wren ask "" --direct ``` ## Semantic layer (MDL) -Git-friendly definition of what data *means*: models, columns, relationships, views, cubes, metrics + `instructions.md` / `queries.yml`. +Wren **и есть** governed semantic layer в **Modeling Definition Language (MDL)** — Git-friendly определение, что данные *значат*, а не только где лежат. + +MDL покрывает: + +- **Models, columns, relationships, views** +- **Cubes and metrics** — approved reusable definitions +- **Business context beyond schema** — enums, units, joins в `instructions.md` и `queries.yml` + +Плюс **AI context layer** — memory, examples, unstructured knowledge — и governed execution engine. + +## Что входит + +- **MDL — semantic layer** +- **Engine**: Apache DataFusion, 22+ sources (BigQuery, Snowflake, PostgreSQL, ClickHouse, Redshift, Databricks, DuckDB, …) +- **GenBI dashboards**: [`wren-core-wasm`](https://docs.getwren.ai/oss/sdk/wasm), Vercel / Cloudflare Pages +- **Knowledge & memory — AI context layer**: `instructions.md`, `queries.yml`, local LanceDB +- **Agent SDK**: `wren-langchain`, `wren-pydantic` +- **Governed execution primitives**: dry-plan, row limits, structured errors + +## Что дальше + +- End-to-end correctness primitives +- Agent-native distribution / SDKs — [Discussions](https://github.com/Canner/WrenAI/discussions) + +Roadmap: [introduction](https://docs.getwren.ai/oss/introduction). + +## FAQ + +### Что такое generative BI (GenBI)? + +BI, который производят AI-агенты: governed SQL, dashboard, share — на AI context layer, чтобы output был trustworthy. + +### Делает ли Wren AI text-to-SQL? + +Да — **governed** text-to-SQL против MDL + dry-plan, и дальше dashboards. + +### Wren AI — это semantic layer? + +Да (MDL) + AI context layer (memory, examples, unstructured knowledge). + +### Что такое AI context layer? + +Reviewable, version-controlled knowledge, которого нет в schema: semantics, definitions, examples, memory, governance. [Vision](https://www.getwren.ai/post/the-missing-context-layer-for-ai-agents-over-business-data). + +### OSS vs Wren AI Cloud / self-hosted? + +OSS в этом репо — free forever, Apache-2.0, self-hostable. RLS/CLS, GenBI UI, GenBI Apps, agentic mode, support/SLA — commercial (Cloud / Enterprise Plus). [Open core boundary](https://www.getwren.ai/en/open-core). + +### Какие data sources? + +22+ через DataFusion: BigQuery, Snowflake, PostgreSQL, ClickHouse, Redshift, Databricks, DuckDB, … [Connect a database](https://docs.getwren.ai/oss/guides/connect). -## What's Included +## Open core: OSS vs Cloud / self-hosted -- MDL semantic layer -- Apache DataFusion engine, 22+ sources -- GenBI dashboards (`wren-core-wasm`) -- Knowledge & memory (LanceDB hybrid retrieval) -- SDKs: `wren-langchain`, `wren-pydantic` -- Governed execution primitives +**Open core.** Context engine (MDL, governed text-to-SQL, MCP, CLI, 22+ connectors) — OSS. Commercial: RLS/CLS, GenBI UI, Scenario AI harnesses, advanced security, support/SLA. [Подробности →](https://www.getwren.ai/en/open-core). -## FAQ (кратко) +## Заметка про имя «GenBI» -- **GenBI** — BI от AI-агентов: governed SQL + dashboards + context. -- **text-to-SQL** — да, governed; дальше dashboards. -- **semantic layer** — да (MDL) + AI context layer. -- **OSS vs Cloud** — OSS free (Apache-2.0); RLS/CLS, GenBI UI — commercial. [Open core](https://www.getwren.ai/en/open-core). -- **Sources** — BigQuery, Snowflake, PostgreSQL, ClickHouse, Redshift, Databricks, DuckDB, … [Connect](https://docs.getwren.ai/oss/guides/connect). +«GenBI» теперь — open-source generative-BI capability. Ранний **Wren AI GenBI** Docker app — **Wren GenBI Classic** на [`legacy/v1`](https://github.com/Canner/WrenAI/tree/legacy/v1). -## License +## Лицензия -Apache-2.0 — see [LICENSE](./LICENSE). +Apache-2.0 — см. [LICENSE](./LICENSE). --- -

⬆ back to top

+

⬆ наверх