gws-shared_skill

This skill helps you standardize authentication, output formatting, and safety checks across the Google Workspace CLI for faster, safer automation.

Rust

19.4k

GitHub Stars

1

Bundled Files

3 weeks ago

Catalog Refreshed

1 month ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstart where the catalogue uses aiagentskills.

npx veilstart add skill googleworkspace/cli --skill gws-shared

SKILL.md2.0 KB

Overview

This skill provides shared CLI patterns for Google Workspace automation, centralizing authentication, global flags, output formatting, and safety rules. It standardizes how agent skills call Drive, Gmail, Calendar, Sheets, Docs, Chat, Admin, and other Google APIs from a single gws command-line tool. The goal is predictable, auditable automation with built-in safeguards for destructive operations and PII handling.

How this skill works

The skill expects the gws binary on PATH and uses either browser-based OAuth or service account credentials via GOOGLE_APPLICATION_CREDENTIALS. Commands follow a single syntax: gws <service> <resource> [sub-resource] <method> [flags]. Global flags control format, dry-run behavior, and sanitization while method flags handle params, body JSON, uploads, pagination, and binary output saving.

When to use it

Automating Google Workspace tasks across multiple services from an agent or CI job
Performing scripted reads and writes while enforcing confirmation for destructive actions
Exporting or transforming API responses into json, table, yaml, csv for downstream processing
Running bulk queries with auto-pagination and controlled page limits/delays
Screening outputs for PII or policy concerns before exposing results

Best practices

Install the gws binary on PATH and verify authentication with gws auth login or service account env var
Use --dry-run to validate commands locally before making changes to live data
Always prompt the user for confirmation before executing write or delete operations
Apply --sanitize with a model template when responses may contain PII or sensitive content
Prefer structured output (--format json) for automation; use table or csv for human-readable exports
Limit auto-pagination with --page-limit and add --page-delay to avoid rate limits

Example use cases

List all Drive files and save results as NDJSON while iterating pages with --page-all
Create a Calendar event via gws calendar events insert --json '{...}' after a dry-run confirmation
Download a Docs export to a file using -o/--output and verify content with --sanitize
Run admin reports using service account credentials in CI and output as csv for analytics
Search Gmail messages with query params and stream results for downstream AI processing

FAQ

Set GOOGLE_APPLICATION_CREDENTIALS to a service account JSON key and ensure the service account has needed scopes.

How can I avoid leaking secrets in outputs?

Never print tokens or API keys. Use --sanitize to screen results and use --dry-run to validate commands without calling APIs.