Populating Google Sheets From Agent Data — Live Dashboards Without Code

The Problem

Every week, someone at PyratzLabs had to compile numbers. Portfolio P&L from Warren's financial reports. Cron job success rates from system logs. Marketing metrics from Peiy's campaigns. Agent uptime scores. All of it needed to land in a spreadsheet for the Friday board review.

That was 1-2 hours every week. Not hard work — just tedious, error-prone, soul-crushing work. Copy from one screen, paste into another. Format the cells. Double-check the numbers. Realize you pulled last week's data for one column. Fix it. Share the link. Repeat next Friday.

The dashboard tools exist — Metabase, Looker, Grafana. But they need a data warehouse, connectors, schemas, maintenance. We're a lean team. We don't need a BI stack. We need a spreadsheet that fills itself.

The Solution

Every agent in our fleet now writes its own operational data directly to Google Sheets using the gog CLI. Warren writes portfolio metrics. The system agent writes cron health data. Peiy writes marketing KPIs. Alfrawd compiles and formats the weekly board sheet every Friday at 8am — before anyone's even asked for it.

The Process

Step 1: Create the sheet structure once

# Create the board KPI spreadsheet
gog sheets create "PyratzLabs — Weekly KPIs" --share "team@pyratzlabs.com"

I set up the headers manually the first time: Date, ARR, MRR Growth, Active Agents, Cron Success Rate, Portfolio Value, Marketing Qualified Leads, NPS Score. That's it. The agents handle everything else.

Step 2: Agent writes data on schedule

# Warren's Friday financial update cron
schedule: "0 7 * * 5"  # Every Friday at 7am UTC
task: |
  Pull this week's financial metrics:
  - Current ARR, MRR, MRR growth %
  - Portfolio total value, weekly change
  Write to Google Sheet "PyratzLabs — Weekly KPIs"
  in the next empty row with today's date.
model: claude-sonnet-4-20250514

# Inside the agent task:
gog sheets append "PyratzLabs — Weekly KPIs" \
  --sheet "Sheet1" \
  --values "2026-03-13,2100000,34%,31,99.2%,€4.2M,127,8.4"

Step 3: Multiple agents, one sheet

# System health agent writes cron stats
gog sheets update "PyratzLabs — Weekly KPIs" \
  --sheet "Agent Health" \
  --range "A2:F2" \
  --values "2026-03-13,31,29,2,99.2%,0"

# Peiy writes marketing data
gog sheets append "PyratzLabs — Weekly KPIs" \
  --sheet "Marketing" \
  --values "2026-03-13,4200,127,31,12,€847"

Step 4: Formatting and chart-ready structure

The agent writes data in a consistent schema every time — same column order, same date format, same number formatting. Google Sheets' built-in charts auto-update because the data range expands predictably. I set up the charts once. They just work.

Step 5: Friday morning compilation

# Alfrawd's Friday morning board prep
schedule: "0 8 * * 5"
task: |
  Read all tabs in "PyratzLabs — Weekly KPIs".
  Generate a 5-line executive summary comparing
  this week to last week. Write to the "Summary" tab.
  Send the summary + sheet link to Bilal on Telegram.

The Results

The best part: no new tools. Everyone already knows Google Sheets. The board members don't need to learn a new dashboard. They get the same spreadsheet they've always seen — it just populates itself now.

Try It Yourself

1. Create a Google Sheet with your KPI structure — headers, tabs, whatever your team expects
2. Configure gog CLI with write access to the sheet
3. Set up one cron job per data source — each agent writes to its own tab
4. Add a compilation cron that summarizes across tabs weekly
5. Share the sheet with your team — they see live data, no training needed

Start with one metric. Get the agent writing a single number to a single cell on schedule. Once that works, expand. Within a week you'll have a self-updating dashboard that cost zero engineering hours to build.

The best BI tool is the one your team already uses. Just remove the human bottleneck.