Actionist·Metabase

Metabase

· #209 most-used

Query, visualise, and act on your data without writing SQL

ProductivityDatabaseAnalyticsDeveloperAutomation

Metabase is an open-source business intelligence platform that lets everyone in your organisation explore and understand data through an intuitive GUI query builder, auto-generated dashboards, and embedded analytics. Connect it to Actionist and your agents can fetch question results, retrieve dashboard data, manage database connections, query metrics, and surface alerts — all without anyone opening the Metabase UI. Whether your team needs a weekly KPI digest, an anomaly alert routed to Slack, or a data export wired straight into a downstream workflow, Actionist brings Metabase's analytics layer into every agent task.

Try Metabase with ActionistVisit Metabase
Average time saved
10 hours
per person · per month
≈ 1 workdays back

Eliminates manual work. Agents eliminate the manual cycle of opening Metabase, running saved questions, copying results into reports, and distributing data digests to stakeholders.

Schedule

What your Metabase agent runs on autopilot

A week of scheduled jobs your Actionist agent will execute on your behalf.

28Scheduled jobs
7Agents at work
24/7Always on
Agents
Tue–Thu
Tue
Wed
Thu
7a
8a
9a
10a
11a
12p
1p
2p
3p
4p
5p
6p
Multi-app workflows

Metabase × every other app you use

End-to-end automations that span multiple apps — each one a real business outcome.

6Workflows
6Apps spanned
~9 hrsSaved / week
6Personas served
For finance
Featured4 apps

Automated weekly financial KPI digest

Every Monday morning, the Finance Agent runs the core financial Metabase questions, writes the results to the P&L tracking sheet, and posts a formatted digest to #finance and #leadership — giving every stakeholder current revenue, expense, and invoice data before the first meeting of the week without a single manual report pull.

~1 hrs

Time saved for your team — every week, on autopilot

The flow
Trigger·Every Monday at 8 AM on a scheduled agent task
Trigger
Step 1
A
Actionist
Scheduled agent task fires at 08:00 Monday
read
Step 2
Metabase
Run 'Weekly Net Revenue', 'Operating Expenses by Category', and 'MRR Movement' questions
read
Step 3
Metabase
Run 'Overdue Invoices by Age Bucket' question
write
Step 4
Google Sheets
Write revenue, expense, and MRR movement figures to the weekly P&L tracking sheet
write
Step 5
Slack
Post formatted KPI digest with variance callouts to #finance and #leadership
Result
Write revenue, expense, and MRR movement figures to the weekly P&L tracking sheetPost formatted KPI digest with variance callouts to #finance and #leadership
The win
Saved per run
45 min
Runs / week
~1×
Finance team enters every Monday with current KPIs before 8:30 AM
Driven byFinance Agent
ROI

Savings

What your team gets back — two angles: what you stop doing manually, and what that's worth.

Without Actionist

What you do manually today

With Actionist

What your agent runs for you

  • Sales
    40 min / week
    Manual pipeline data pulls before calls

    Reps open Metabase, navigate to the pipeline dashboard, screenshot the numbers they need, and paste them into the pre-call prep doc — 10 minutes per rep before each major call.

    Sales Agent
    0 min
    Agent delivers live pipeline data before every call

    The Sales Agent runs the 'Open Pipeline by Stage' question every Monday and on demand, posting results directly to Slack so reps have current numbers without ever opening Metabase.

  • Marketing
    45 min / week
    Weekly manual campaign metrics assembly

    The marketing team pulls acquisition and conversion data from Metabase each Tuesday, copies numbers into a Google Sheet, and formats a summary — taking 45 minutes before every planning meeting.

    Marketing Agent
    0 min
    Agent assembles and posts campaign metrics automatically

    The Marketing Agent runs Metabase questions twice a week, writes results to the reporting sheet, and posts a digest to #marketing — the sheet is always current and the meeting always starts with real numbers.

  • Customer Support
    75 min / week
    Daily manual ticket volume check in Metabase

    The support lead opens Metabase every morning to check ticket volume by category, compares it mentally to yesterday, and decides whether to escalate — a 15-minute ritual before the first ticket response.

    Customer Support Agent
    0 min
    Agent detects anomalies and alerts the team before the queue backs up

    The Support Agent runs daily volume queries, compares to the rolling average, and alerts #support-ops only when a category spikes above 2x normal — the team is notified before the queue backs up, not after.

  • Human Resources
    30 min / week
    Manual headcount and hiring funnel report assembly

    The HRBP queries Metabase for headcount and hiring funnel data each Monday, copies the numbers into a slide deck, and shares it before the people review — 30 minutes of data assembly before every weekly sync.

    Human Resources Agent
    0 min
    Agent delivers headcount and hiring metrics before the review

    The HR Agent runs Metabase questions Monday and Wednesday, writes results to the people ops sheet, and posts summaries to #people-ops — the HRBP enters every review with current data already formatted.

  • Finance
    45 min / week
    Manual weekly financial report extraction

    Finance runs Metabase revenue and expense questions each Monday, copies the figures into the P&L sheet, and formats the weekly digest — a 45-minute process that delays the board dashboard refresh every week.

    Finance Agent
    0 min
    Agent populates the P&L sheet and posts the digest automatically

    The Finance Agent runs Metabase financial questions at 8 AM Monday, writes figures to the P&L sheet, and posts the digest to #finance and #leadership — board data is current before the first meeting of the week.

  • Operations
    30 min / week
    Manual Metabase hygiene audits

    The data ops team manually reviews question lists, alert configurations, and database connections monthly — a 2-hour audit that often reveals stale reports and monitoring gaps that have been silently broken for weeks.

    Operations Agent
    0 min
    Agent audits questions, alerts, and databases every week

    The Operations Agent lists all Metabase questions, alerts, and databases weekly, flags stale reports and monitoring gaps, and posts findings to #data-ops — issues are caught in days, not months.

  • Legal
    60 min / week
    Manual weekly compliance monitoring check

    The legal team manually reviews the Metabase alert list, cross-references it against the monitoring matrix, and checks the database inventory against the data processing register — a 60-minute weekly compliance ritual.

    Legal Agent
    0 min
    Agent verifies monitoring coverage and database compliance automatically

    The Legal Agent compares all Metabase alerts to the monitoring matrix and all databases to the approved register every Monday, posts gaps to #legal-ops, and appends coverage to the DPO's weekly log — in under 10 minutes.

+ 100s of other Metabase automations
Average time saved
33 hrs / person / month
Calculator

Calculate what your team saves

Team size
5 people
Hourly rate
$75 / hr
Hours saved / week
13
Hours saved / year
625
Annual ROI
$46,875

Based on Metabase's typical team usage — the visible tasks plus a few other automations the agent runs: ~2.5 hrs / person / week of admin work automated.

Connect

How to plug Metabase into Actionist

Pick the connection method that suits your environment.

Generate a Metabase session token by authenticating via the `/api/session` endpoint with your admin credentials, then paste the token into Actionist. For long-running connections use a dedicated service account with read and write permissions scoped to the databases your agents need.

1
Open Metabase Admin Settings

Log in to Metabase as an admin, go to Settings → Admin → API keys (available in Metabase v0.48+). Click 'Create API key' and give it a descriptive name like 'Actionist Agent'.

2
Copy your API key

Metabase will show the key once — copy it immediately and store it in a secure location. This key inherits the permissions of the group you assign it to.

3
Paste into Actionist

Enter your Metabase instance URL and paste the API key into the fields below. Actionist will run a test call to /api/user/current to verify the connection.

Credentials you'll need
Metabase URL*
Your Metabase instance URL, e.g. https://analytics.yourcompany.com
API Session Token*
POST /api/session with {username, password} to get your session token, or generate a long-lived API key in Metabase Settings → Admin → API keys (Metabase 48+).
Actions

14 actions your agent can call

Read and write operations available to your Actionist agent.

Triggers

0 events your agent can react to

Events your agent watches for, and the actions it kicks off in response.

This app has no triggers yet.
MCP servers

MCP servers that work with Metabase

Connect Actionist to MCP servers built for or around this app.

metabase-mcp

MCP server connecting Claude to Metabase with 28 tools for natural language data analysis, dashboard management, SQL queries, and automation.

metabase-ai-assistant

MCP Server for Metabase with 111 tools for SQL generation, dashboard management, and PostgreSQL integration.

Metabase
Official

Official Metabase MCP server (released with Metabase 60, April 2026) that lets AI clients search, explore, query, and visualize data in a Metabase instance.

FAQs

Questions about Metabase + Actionist

How does Actionist connect to Metabase?
Go to the Apps tab in Actionist, find Metabase, and click Connect. You'll need two things: the URL of your Metabase instance (e.g. https://analytics.yourcompany.com) and an API key. In Metabase v0.48 and later, generate a long-lived API key in Settings → Admin → API keys — create one with the permissions of a service account group scoped to the databases your agents need. For older Metabase versions, you can generate a session token by POSTing your credentials to /api/session. Paste either into Actionist and it will run a test call to /api/user/current to confirm the connection before any agent actions run.
Do Actionist agents need admin permissions on Metabase?
No — and they shouldn't have them. Create a dedicated Metabase group for the Actionist service account with read access to the question collections and databases your agents need, plus write access only to the specific resources (like creating questions) you want to enable. Most agent tasks — running questions, fetching metrics, listing alerts, querying databases — only require read permissions. Add write permissions (Add Database, Create Question) selectively to the agent group. Principle of least privilege applies here exactly as it does with any API integration.
Can Actionist agents run live queries against my database through Metabase?
Yes, via two paths. First, the agent can run Get Question Result Data to execute any saved Metabase question and return its current results — this uses the same query engine as the Metabase UI, respecting your row limits and caching settings. Second, the agent can run Run Ad-hoc Query to execute a native SQL or MBQL query directly against any connected Metabase database, without needing a pre-saved question. Both paths respect the permissions of the Metabase service account — the agent can only query databases and tables the account has access to.
What are the most common things Actionist agents do with Metabase?
The four patterns that come up most often: (1) scheduled KPI digests — running saved questions on a weekly cadence and posting results to Slack or Google Sheets so teams start meetings with current data; (2) on-demand data lookups — letting team members ask data questions in Slack that the agent answers by querying Metabase in under a minute; (3) anomaly detection — running a Metabase question daily and comparing the result to a threshold or rolling average, then alerting the team when something looks off; (4) instance hygiene — regularly listing stale questions, orphaned alerts, and unregistered databases to keep the Metabase instance clean and auditable.
Can Actionist work with self-hosted (open-source) Metabase as well as Metabase Cloud?
Yes. Actionist connects to Metabase via the REST API, which is identical on both self-hosted (open-source and Enterprise) and Metabase Cloud instances. You just need a network-accessible URL for your instance and a valid API key or session token. If your self-hosted Metabase is behind a VPN or firewall, ensure that Actionist's outbound IP range is allowlisted on your network so the agent can reach the /api endpoints.
Does Actionist support the new Metabase MCP server?
Yes. Metabase released an official MCP server with Metabase 60 (April 2026) that exposes search, exploration, query, and visualization capabilities through the Model Context Protocol. Actionist supports MCP connections, so you can connect to Metabase using the official io.github.metabase--mcp server as an alternative to the API key method. The MCP path gives agents access to a broader set of Metabase tools including natural language querying via Metabot's semantic layer. Choose the connection method that best fits your Metabase version and security posture.
Can Actionist agents create new Metabase questions programmatically?
Yes, via the Create Question action. The agent can define a new saved question using MBQL (Metabase Query Language) or native SQL, specify the display type (table, bar chart, number, etc.), and assign it to a collection — all without anyone opening the Metabase editor. This is useful for standardising reporting when a new feature ships, creating client-scoped reports on onboarding, or replicating a question template for a new region. Note that the service account must have create permissions on the target collection.
How does Actionist handle Metabase's row limits and query caching?
When the agent runs Get Question Result Data, Metabase applies the same row limit that's configured for that question in the Metabase UI (typically 2,000 rows by default for GUI questions; configurable per instance). The result also reflects Metabase's caching settings — if the question result is cached, the agent gets the cached data, which may be up to the cache TTL old. For the freshest data, use Run Ad-hoc Query, which bypasses the cache and queries the database directly. If you need more than 2,000 rows in a result, contact your Metabase admin to adjust the row limit for the relevant question.