GA4 MCP for Pi: Setup, Skills and First Query

angle: One exact Pi mcp.json block for GA4, so the agent-dir file stops getting confused with Claude paths.

I query GA4 from inside Pi with the same server I run in OpenCode. GA4 MCP for Pi gives you that: one entry in the agent-dir mcp.json, your property ID plus credentials in env, then plain questions answered from your own property schema.

Proof strip

Configuring Pi to launch the server

Prerequisites: numeric GA4_PROPERTY_ID (Admin, Property details, not the G- ID), plus a service-account JSON path or gcloud auth application-default login. Keep the JSON path absolute.

1. Open your agent-dir mcp.json and add:

{
  "mcpServers": {
    "google-analytics": {
      "command": "uvx",
      "args": ["google-analytics-mcp"],
      "env": {
        "GA4_PROPERTY_ID": "123456789",
        "GOOGLE_APPLICATION_CREDENTIALS": "/absolute/path/to/key.json"
      }
    }
  }
}

A root-level server map without the mcpServers wrapper is also accepted; I show the wrapped shape because it ports cleanly from Claude configs.

2. For stdio servers, Pi passes a small safe base plus allow-listed vars (MY_PI_CHILD_ENV_ALLOWLIST / MY_PI_MCP_ENV_ALLOWLIST) plus explicit env above. Prefer explicit env for the two GA4 values.

3. Open /codemcp in Pi to confirm the server, or restart Pi and check the tool list.

4. Verify by asking: "List my GA4 properties." You have succeeded when list_properties returns rows with no setup error.

Skills worth loading first

| Skill slug | What you use it for in Pi |

|---|---|

| traffic-diagnosis | You saw sessions move and want the ordered checks |

| common-metric-names | Your field name failed and you want the working GA4 name |

| channel-acquisition | You want sessions split by source, medium, channel group |

| compatible-combinations | Your dimensions and metrics returned a 400 incompatibility |

| date-ranges | You compare two periods and want the two-query pattern |

Full set of 15 at https://ga4mcp.com/skills/, loaded with search_skills("<slug>").

Running your first query

Ask: "What were my top channels last week?"

Your agent should call:

get_ga4_data(
  dimensions=["sessionDefaultChannelGroup"],
  metrics=["sessions"],
  date_range_start="7daysAgo",
  date_range_end="yesterday",
  intent="which channels drove sessions last week"
)

Read the period figure from totals. GA4 computes that block; hand sums drift and I stopped doing them.

Fixing the three errors you will hit

1. Server missing after edit. Symptom: /codemcp omits google-analytics. Fix: confirm the file is the agent-dir mcp.json Pi reads, valid JSON, then restart Pi and trust the project where prompted.

2. Setup error on first call. Symptom: missing property or credentials. Fix: confirm both env values with an absolute key path and allow-list coverage; for ADC expiry run gcloud auth application-default login. setup_ga4_access recovers mid-session where prompts are supported.

3. Invalid dimension, metric, or filter. Symptom: 400 naming error. Fix: run search_schema("<keyword>"), load common-metric-names plus filter-structures, retry. Offline guide: get_troubleshooting_guide(topic="schema").

FAQ

Does GA4 MCP for Pi use mcp.json?

Yes. Agent-dir mcp.json with the mcpServers block above, verified in pi-codemcp docs today.

Do I need a Pi extension for MCP?

No for the server entry. pi-mcp and pi-codemcp add discovery and code-mode surfaces; the block above is the wiring either way.

Does the installer cover Pi?

The site auto-configure list does not name Pi. I show the hand block because it is exact and auditable.

Why does the agent guess wrong field names?

Training predates GA4 renames (UA sunset 2023-07-01; conversions renamed to key events 2024-05-06). The ua-to-ga4 skill maps each old name.

Is usage read-only and private?

Yes. Tools read metadata and reports; they change nothing. Traffic runs from your machine to the GA4 Data API. Anonymous counts stop with DISABLE_TELEMETRY=1.

Why does the repo read 2.11.5 when the site says 2.11.4?

Main moved ahead of the site label at fetch today. I quote 2.11.4 for installs and state the gap here.

Freshness and machine links

Verified 2026-10-09 against https://ga4mcp.com/ (v2.11.4), https://ga4mcp.com/setup/, https://ga4mcp.com/skills/, Pi package plus pi-codemcp MCP docs, and repo raw pyproject.toml 2.11.5.

Version gap (2026-10-09): repo pyproject.toml reads 2.11.5 [source: repo pyproject.toml, verified this run] while the live site labels v2.11.4 [source: https://ga4mcp.com/ fetch 2026-10-09]; installs on this page quote the site label.

Machine: https://ga4mcp.com/llms.txt, https://ga4mcp.com/data/tools.json, https://ga4mcp.com/data/skills.json.

Breadcrumb

Hub (/ga4-mcp-for-every-harness/) / GA4 MCP for Pi (you are here).