{"slug": "claude-dashboards-you-can-govern", "title": "Claude Dashboards you can Govern", "summary": "Dbt Labs launched a dbt Charts connector for Claude, ChatGPT, Codex and any other agent tool that speaks MCP, returning dashboards drawn in the conversation, run on the customer's warehouse and stored as a single YAML file in the repo. Every call is checked against the asking user's own permissions, and with branch mode enabled each saved dashboard becomes a pull request reviewed before going live. On dbt Labs' BoardBench benchmark, across the last six rounds of about 160 runs per side on 14 held-out datasets, Codex wrote 7,800 output tokens with dbt Charts versus 16,600 building a dashboard its own way.", "body_md": "Claude can build dashboards now. Ask for one and it appears in the conversation a minute later, and anyone at your company can do it.\n\nThe question is what happens next. The dashboard lives in one person’s chat. Nobody reviewed the query, nobody checked whether that person should see the numbers, and the data team hears about it when a figure from it is quoted in a meeting. The usual answer is to treat the chat as a sketchpad: prototype the dashboard there, then rebuild it in a BI tool before anyone relies on it.\n\nSelf-serve analytics finally arrived. It shouldn’t need a rebuild to be trusted. The dbt Charts connector gives you both: anyone can ask for a dashboard, and the data team still owns every number in it.\n\n## Dashboards as (governable) code\n\nToday we’re launching the dbt Charts connector for Claude, ChatGPT, Codex, and\nany other agent tool that speaks MCP. Add it to the assistant your company\nalready uses, and the same request returns a dashboard written in\n[dbt Charts](https://dbtcharts.com/blog/charts-built-for-chat/): drawn in the conversation, running\non your warehouse, and stored as one YAML file a person can read.\n\n**For the person asking**\n\n- Ask in the assistant you already use: Claude, ChatGPT, Codex, and anything else that speaks MCP.\n- Get a live dashboard in a minute, with no SQL, no install, and no ticket.\n\n**For the data team**\n\n- Every call is checked against that person’s own permissions.\n- Every saved dashboard is a file in your repo. Turn on branch mode and each one is a pull request, reviewed before it goes live.\n- Nothing to rebuild, and nothing ungoverned to chase down.\n\nThe person asking sees a chart, and never YAML or SQL. They can ask for changes, click into the editor to adjust a label, or ask the assistant to save it to a folder. The data team sees a file in the repo.\n\n## Less for the data team to review\n\nGovernance only works if someone can read what the assistant built.\n\nBoardBench is our benchmark for agent-built dashboards. An agent gets a written brief and a dataset, builds a dashboard, then handles a styling request and a set of review notes. Left alone, it writes HTML, CSS, JavaScript, and a script to prepare the data. In dbt Charts it is one YAML file. Across our last six rounds, about 160 runs a side on 14 held-out datasets, Codex wrote 7,800 output tokens with dbt Charts and 16,600 building a dashboard its own way: less than half as much for the data team to read.\n\n## And it gets more answers right\n\nSelf-serve only works if the answers are right. Give the assistant the data team’s project and it stops guessing.\n\nAskBench is our benchmark for data questions. An agent gets a question a real team asked of its warehouse and has to come back with the right number. We run each model twice: once with only the warehouse, and once with a dbt Charts project in reach.\n\n## See the YAMLa example\n\n```\ncharts:\n  right_answers:\n    query: askbench\n    type: bar\n    title: \"AskBench: right answers with and without dbt Charts\"\n    subtitle: Percent of real data questions answered right. Gray is the agent on the bare warehouse; orange is the same model with a dbt Charts project in reach.\n    x: model\n    y: with_charts\n    y_start: bare\n    style:\n      orientation: horizontal\n      legend:\n        visible: false\n      marks:\n        bar:\n          band_width: 0.03\n          color: \"#b5b5b5\"\n      axis_y:\n        grid:\n          torn_baseline:\n            enabled: true\n        scale:\n          continuous:\n            domain: [30, 100]\n    layers:\n      - type: scatter\n        y: bare\n        label: Bare warehouse\n        style:\n          marks:\n            point:\n              size: 140\n              color: \"#9a9a9a\"\n      - type: scatter\n        y: with_charts\n        label: With dbt Charts\n        style:\n          marks:\n            point:\n              size: 140\n              color: \"#FE6703\"\n              labels:\n                visible: true\n                field: gain\n                position: right\n                dx: 10\n```\n\nWe ran six Claude and OpenAI models on the same 197 questions, and every one got more answers right with the project. Claude Opus went from 61% to 89%, Claude Haiku from 37% to 78%, and GPT-6.1 Sol from 81% to 94%. The agent stops guessing at which table holds revenue and reads the query a dashboard already runs.\n\nThe warehouse does less work too: on GPT-6 Sol the project scanned 37% fewer bytes.\n\n## Saved next to your dbt models\n\nA dashboard sketched in a chat and rebuilt in a BI tool is two dashboards: the one someone asked for and the one the data team maintains. They drift apart, and the second lives in a vendor’s format that only that vendor’s product can edit.\n\nWith dbt Charts there is one. The dashboard the assistant drew is a YAML file in your dbt project, beside the models it reads:\n\n```\nmy_dbt_project/\n  dbt_project.yml\n  profiles.yml\n  models/\n  charts/\n    sales/\n      bookings_and_recurring_revenue.yml\n      top_salespeople.yml\n    themes/\n      newsroom.yml\n```\n\nA person can read it, a reviewer can see what changed, and a renamed column fails the pull request that renames it. dbt Charts is open source, so if you ever stop using it, you keep a folder of readable files and the SQL inside them.\n\n## What the data team keeps\n\n- **It signs in as you.** There is no service account. Every call is checked\nagainst that person’s own access to the dashboard, the project, and the data\nsource, so the assistant can reach at most what the person already can.\n- **It carries queries, not data.** The file holds the SQL, so it runs against\nthe warehouse each time it opens, and each query is labeled in your\nwarehouse’s job history.\n- **It saves only when asked, and it saves to Git.** What the assistant builds\nis a private draft. Saving it is a commit in your repo, and with[branch mode](https://dbtcharts.com/platform/#visual-prs) on, a pull request reviewed as a picture\nbefore it goes live.\n- **It can be switched off.** Account Settings lists every assistant with\naccess. Disconnect one and its next call fails.\n\n## Add it to your tools\n\nThe connector is a remote server at one URL, so there is nothing to install. On Claude’s Team and Enterprise plans an owner adds it once and everyone in the organization gets a Connect button.\n\n- **Claude** , on the web, desktop, and mobile:[add the connector](https://dbtcharts.com/platform/mcp/#connect) .\n- **ChatGPT** :[add the connector](https://dbtcharts.com/platform/mcp/#connect) by URL and sign in.\n- **Codex** :`codex mcp add dbt-charts --url https://dbtcharts.com/mcp` .\n- **Claude Code and Cowork** : the[dbt Charts plugin](https://dbtcharts.com/platform/mcp/#connect) bundles the connector with the\nskills for building, styling, and reviewing dashboards.\n- **Anything else that speaks MCP** : point it at`https://dbtcharts.com/mcp` and sign in.\n\nEvery tool and permission is listed in the\n[connector docs](https://docs.dbtcharts.com/platform/claude-chatgpt-connector/).\n\n## Try it\n\nYour business users get answers in the tools they already use. Your data team keeps the models, the reviews, and the final say.\n\nThe connector is free to add, and it works with the free workspace. There’s no license to buy and no demo to book.\n\n- Connect: [dbtCharts.com/platform/mcp](https://dbtcharts.com/platform/mcp/)\n- Docs: [the connector guide](https://docs.dbtcharts.com/platform/claude-chatgpt-connector/)\n- Hosted BI: [dbtCharts.com](https://dbtcharts.com)\n- Community: [#dbt-charts](https://getdbt.slack.com/archives/C0C1SFL0QN5) on the[dbt Community Slack](https://www.getdbt.com/community/join-the-community)\n\nAdd it, ask for a dashboard, and tell us what it got wrong.", "url": "https://wpnews.pro/news/claude-dashboards-you-can-govern", "canonical_source": "https://dbtcharts.com/blog/claude-dashboards-you-can-govern/", "published_at": "2026-11-17 00:00:00+00:00", "updated_at": "2026-11-17 20:01:54.422131+00:00", "lang": "en", "topics": ["ai-agents", "agent-protocols", "ai-tools", "developer-tools", "structured-data"], "entities": ["dbt Labs", "dbt Charts", "Claude", "ChatGPT", "Codex", "MCP", "BoardBench", "AskBench"], "also_reported_by": [], "takeaway": "Adding the dbt Charts MCP connector to an assistant turns dashboard requests into a single reviewable YAML file, and on BoardBench Codex used 7,800 output tokens with dbt Charts versus 16,600 building the dashboard its own way.", "recommendation": null, "alternates": {"html": "https://wpnews.pro/news/claude-dashboards-you-can-govern", "markdown": "https://wpnews.pro/news/claude-dashboards-you-can-govern.md", "text": "https://wpnews.pro/news/claude-dashboards-you-can-govern.txt", "jsonld": "https://wpnews.pro/news/claude-dashboards-you-can-govern.jsonld"}}