Google Workspace CLI — one command-line tool for Drive, Gmail, Calendar, Sheets, Docs, Chat, Admin, and more. Dynamically built from Google Discovery Service. Includes AI agent skills.
29.6k
Stars
1.7k
Forks
128
Open issues
30
Contributors
AI Analysis
gws is a unified CLI for Google Workspace (Drive, Gmail, Calendar, Sheets, Docs, Chat, Admin APIs) written in Rust, dynamically built from Google's Discovery Service. It serves developers and AI agents who need programmatic access to Workspace APIs without boilerplate, offering structured JSON output and 40+ agent skills. It is primarily for technical users comfortable with CLI tools and AI integrations, not for end users seeking GUI-based Workspace access.
Inferred from signals mentioned in the README (tests, CI, type safety) — not a review of the actual code.
AI's overall editorial judgment — not an average of the bars above, can weigh other factors too.
Google Workspace CLI dynamically built from Discovery API, with AI agent skills baked in
gws is a Rust-based CLI that exposes the entire Google Workspace API surface — Drive, Gmail, Calendar, Sheets, Docs, Chat, Admin, and more — through a single unified tool. Its defining technical choice is generating its command surface dynamically from Google's Discovery Service, meaning it tracks API additions automatically. It also includes 40+ agent skills targeting LLM/AI agent workflows. Built for developers, DevOps engineers, and AI agent pipelines that need scriptable, structured JSON access to Workspace without writing custom REST integrations. With 27K+ stars in roughly 3.5 months, it has captured significant developer attention quickly.
Created March 2, 2026, under the googleworkspace GitHub organization — notably marked 'not an officially supported Google product' despite the org namespace. Reached 27K stars by June 2026, suggesting strong organic traction from Google-adjacent developers and AI tooling enthusiasts.
The combination of an official-looking Google org namespace, a compelling technical premise (Discovery-driven dynamic CLI), and the AI agent skills angle likely drove viral sharing in developer and AI engineering communities. The repo launched into a moment of intense interest in LLM tool use and MCP-style agent tooling, which appears to have amplified adoption signals. 246 stars in the last 7 days suggests sustained but moderating momentum from an early spike.
Adoption not verified in production environments. The star count is high for a 3.5-month-old repo, but stars do not confirm production deployment. No case studies, testimonials, or production usage reports are referenced in the README. npm package exists (@googleworkspace/cli), which provides a download vector but download counts are not cited. The AI agent skills angle may be driving exploratory rather than production use at this stage.
Written in Rust, distributed as a pre-built binary or via npm (which downloads the binary), Cargo, Homebrew, or Nix. The core architectural decision — dynamically building commands at runtime from Google's Discovery Service JSON — appears well-suited to the problem of keeping pace with a large, evolving API surface. Auth layer reportedly uses AES-256-GCM encryption at rest with OS keyring integration. Likely parses Discovery API schemas to generate CLI argument structures, help text, and JSON schema introspection via `gws schema`. Pre-v1.0 with explicitly warned breaking changes.
CI badge present (GitHub Actions workflow on main branch). Specific test coverage metrics are not documented in README.
Last push June 10, 2026 — 14 days before evaluation date. Active CI pipeline visible. Project explicitly self-describes as under active development pre-v1.0. Maintenance appears active given the recency of commits and stated development trajectory.
ADOPT IF: you need scriptable, structured access to Google Workspace APIs in CI pipelines, agent workflows, or admin automation and don't want to maintain custom REST/OAuth integrations. AVOID IF: you need production stability guarantees — the project is explicitly pre-v1.0 with warned breaking changes, and it requires setting up a Google Cloud project with OAuth credentials, which adds setup friction for casual use. MONITOR IF: you're building LLM agent tooling that needs Workspace integration — the 40+ agent skills and structured JSON output are well-aligned but the project needs to stabilize before depending on it in critical workflows.
Independent dimensions
Mainstream potential
7/10
Technical importance
8/10
Adoption evidence
3/10
- Pre-v1.0 with explicitly warned breaking changes — early adopters in automated pipelines risk regressions without notice.
- Requires a Google Cloud project and OAuth app setup, including scope limitations for unverified apps (85+ scopes fail for testing-mode OAuth) — onboarding friction is non-trivial for non-GCP-familiar users.
- Dynamic command generation from Discovery Service means a Google API change or Discovery schema update could silently break existing workflows.
- Despite the googleworkspace org namespace, this is explicitly not an officially supported Google product — long-term maintenance continuity depends on community and individual contributors, not Google SLAs.
- The AI agent skills angle competes with fast-moving MCP and tool-use standards; if the ecosystem converges on a different integration model, the agent skills layer may need significant rework.
Likely to become a standard-issue tool in developer and DevOps Workspace automation workflows if it reaches v1.0 with stable semantics. The AI agent skills angle may attract significant enterprise interest but requires API stability first.
Newsletter
Get analyses like this every Monday
Free weekly digest of the most interesting open-source discoveries.
Languages
Information
- Language
- Rust
- License
- Apache-2.0
- Last updated
- 1w ago
- Created
- 4mo ago
- Analyzed with
- anthropic/claude-haiku-4-5
Stars over time
Contributors over time
Top 100 contributors only — repos with more will plateau at 100.
Open issues
Bug: raw calendar.events.list fails with insufficient scopes; +agenda helper succeeds with the same account
Docs: no supported way to build a Gmail web link from an API ID (inverse of #790)
GOOGLE_WORKSPACE_CLI_CREDENTIALS_FILE ignored when credentials.enc exists (inverts documented precedence)
Feature request: docs +insert-image helper to embed Drive-hosted images without requiring public sharing
Access blocked: gws has not completed the Google verification process
Top contributors
Similar repos
taylorwilsdon/google_workspace_mcp
This is a comprehensive MCP (Model Context Protocol) server that enables AI...
googleworkspace/python-samples
Official Python sample code repository for Google Workspace APIs, providing...
| Repository | Stars | Week Δ | Language | Score | Updated |
|---|---|---|---|---|---|
|
|
29.6k | +239 | Rust | 8/10 | 1w ago |
|
|
8.1k | — | Go | 8/10 | 23h ago |
|
|
2.8k | — | Python | 8/10 | 3d ago |
|
|
1.4k | — | Python | 7/10 | 2w ago |
|
|
5k | — | Python | 7/10 | 1w ago |
|
|
5.2k | — | JavaScript | 7/10 | 2w ago |
A Go-based Google Workspace CLI with 7,793 stars, predating gws. Less stars but potentially more battle-tested. gws differentiates on dynamic Discovery-based command generation vs. likely static command definitions, and on Rust vs. Go. Different tradeoffs around startup time, ecosystem, and maintenance model.
A Python-based MCP server for Workspace targeting AI agent use cases specifically. gws overlaps heavily on the AI agent angle but is broader (human CLI use too) and distribution-model differs (binary vs. Python runtime). MCP-native tooling may be preferred by Claude/MCP-first agent pipelines.
Google's own agents-cli (3K stars, Python) targets AI agent tooling more broadly. Less focused on Workspace specifically. Likely complements rather than competes directly with gws, but both address LLM tool use patterns.
Gemini CLI (105K stars) is a conversational AI shell, not a Workspace API CLI. Overlap exists in agent-skill territory and both live in Google's GitHub orgs, but they address different primary use cases. Gemini CLI's massive star count reflects a different category entirely.
The official Google Cloud SDK covers some Workspace admin tasks but is primarily GCP-focused. gws targets Workspace productivity APIs (Drive, Gmail, Sheets, etc.) that gcloud largely ignores. They are complementary tools for different API families.