Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Google Sheets Rube Playbook

    1

    Deterministic Google Sheets automation through Rube MCP: read, append, upsert, filter, and format with verified read-backs, not guessed tool calls.

    Free

    5 usesSecurity scanned
    Google Sheets Rube Playbook

    Google Sheets Rube Playbook

    Example session with this skill installed

    Pull the rows from our Q3 Lead Tracker (https://docs.google.com/spreadsheets/d/1AbCdEfGhIjKlMnOp/edit#gid=123456) and sync them into the CRM Sync tab of the ops workbook, matching on the Email column. New leads should be appended; existing ones updated with the latest Stage and Owner. After the sync, freeze and bold the header row and set a basic filter showing only Stage equals Open. Do not delete anything — if rows look stale, flag them and report exactly what changed.

    • Read your context and instructions
    • Compiled the google sheets rube

    Summary

    Synced 42 leads from Q3 Lead Tracker into the CRM Sync tab, matching on the Email column: 29 existing rows updated and 13 new rows appended. No rows were deleted; readback confirmed the header, key column, and final row count.

    MetricBeforeAfter
    Rows in CRM Sync87100
    Rows updated in place—29
    New rows appended—13
    Stage set to Open (visible in filter)4152
    Header row styled/frozenplainbold + frozen

    Next steps

    • Review the 13 new leads now owned by Ravi
    • 4 rows flagged as stale (no activity since June) — confirm before clearing
    • Filters hide Closed-Opportunities; remove or widen the basic filter before any export
    • Want this as a weekly job? I can batch the sync to stay under Sheets quota

    Connects securely to your tools. The creator never sees your data.

    About this skill

    The problem

    Agents often struggle with spreadsheet automation, relying on fragile browser clicks or guessing cell ranges. This leads to data corruption, missing rows, and failed syncs when tab names or structures change.

    What it does

    • Performs deterministic reads and writes using Rube MCP tools instead of simulated UI actions.
    • Resolves spreadsheet and worksheet identifiers dynamically to prevent "tab not found" errors.
    • Executes complex data operations including keyed upserts, batch updates, and row appending.
    • Applies structural changes and styling like freezing panes, header formatting, and basic filters.
    • Implements a verify-then-write loop with mandatory read-backs to ensure data integrity.

    Frameworks & tools

    Google Sheets API, Rube MCP, OAuth 2.0.

    Why this beats prompting it yourself

    Manual prompting usually fails at handling A1 notation syntax and the distinction between spreadsheet IDs and worksheet GIDs. This skill enforces a strict preflight sequence that validates tool schemas and connection states before a single cell is touched.

    Use cases

    • Syncing CRM lead data into a shared tracking sheet with deduplication logic.
    • Automating weekly inventory logs by appending records to the first empty row.
    • Formatting raw data exports into client-ready reports with bold headers and filters.
    • Searching large datasets for specific SKUs or customer emails across multiple tabs.

    Known limitations

    Subject to Google Sheets API per-minute rate limits. Large datasets require chunked reading as unbounded ranges may timeout.

    Depends on Rube's currently exposed tool schema, so exact parameter names may drift between versions. It cannot complete the Google OAuth handshake itself — that stays a one-time user action.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    5 uses

    Times used to date

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 2 months ago

    • Free with an Agensi account

    Requires

    Googleapis

    Permissions

    Network Access

    Allowed Hosts

    sheets.googleapis.com

    Network Access is needed solely for the Rube MCP bridge to call Google Sheets endpoints authorized by your own OAuth connection; no local file access is required.

    Listed2 months ago

    What's inside

    Frequently Asked Questions

    This skill utilizes Rube MCP to perform deterministic API calls rather than fragile browser simulations, ensuring exact data placement and preventing the common 'tab not found' or range errors seen with manual prompting.

    The skill is compatible with any AI agent or platform that supports the Model Context Protocol (MCP) and has the Google Sheets API enabled.

    Yes, the playbook includes logic for keyed upserts and deduplication, allowing the agent to update existing rows based on a unique identifier instead of just appending new data.

    You will receive the complete Playbook configuration, including the specific tool definitions, preflight validation sequences, and the Rube MCP integration schema.

    You must have an active Google Cloud project with the Google Sheets API enabled and valid OAuth 2.0 credentials to allow the Rube MCP server to communicate with your spreadsheets.

    While highly efficient, the skill is subject to Google's API rate limits; for very large datasets, the playbook is designed to use chunked reading to prevent timeouts and ensure stability.

    No. The skill drives Google Sheets through the Rube MCP bridge; you authorize the Sheets toolkit once via OAuth and the playbook handles everything else.

    It resolves the spreadsheet ID, tab name, and a bounded A1 range before any write, chooses the least destructive mode (append, update, upsert, or clear), and verifies each important edit with a readback.

    The included preflight searches for the currently exposed Sheets tools and reads their live schemas, so the agent adapts to naming changes instead of relying on hard-coded examples.

    New in LLM Ops