Google opens a dedicated API for grounding AI agents in its developer docs
In short: Google has released the Developer Knowledge API, a structured, machine-readable source for Google Cloud, Firebase and Android documentation aimed at AI agents and developer tooling. Instead of scraping web pages or relying on an LLM's training cutoff, agents and CLIs can now query fresh, Markdown-formatted docs through several interfaces: a gcloud CLI surface, an installable agent skill, client libraries, and a browser-based API explorer. The underlying index is refreshed frequently so retrieved answers stay close to the latest published documentation.
This summary was generated automatically by AI from Google Developers's publication. It is our own text, not a copy of the original — facts, figures and quotes belong to the source, linked above and below.
What changed?
- 1gcloud CLI gains answer-query, documents search-chunks and documents describe commands, pre-installed in Cloud Shell and available on Linux, macOS and Windows
- 2An agent skill (installed via 'npx skills add google/skills --skill retrieving-developer-knowledge') connects coding assistants such as Antigravity, Claude Code, Cursor and GitHub Copilot through standard MCP, falling back to a REST/curl call when MCP isn't available
- 3Client libraries ship for seven languages: C#, Go, Java, Node.js/TypeScript, PHP, Python and Ruby, exposing AnswerQuery, SearchDocumentChunks, GetDocument and BatchGetDocuments
- 4BatchGetDocuments lets a single call fetch as many as 20 documents at once for efficiency in CI/CD or knowledge-bot pipelines
- 5An interactive APIs Explorer lets developers try the core endpoints without writing any code
Why it matters
For teams that build or operate AI agents, coding assistants, or automated workflows on Google Cloud, this gives a sanctioned way to keep agent responses current without maintaining custom scrapers, and the MCP-based skill means popular coding assistants can plug in with minimal setup.
What it means for AI agents and contact centers
If your company builds voice or coding agents that call Google Cloud or Firebase services, this API could be wired in via MCP or a client library so an agent can pull accurate, up-to-date service documentation (e.g., for release notes or configuration details) during a conversation or a build step, rather than risking outdated or hallucinated answers.
🧪 Worth Testing
Worth testing for any agent stack that already supports MCP or makes tool calls, to see whether grounding responses in live Google documentation reduces errors and how much latency/token overhead the retrieval step adds.
Sources
- Google DevelopersOfficialPrimary sourceOriginal article →„Supercharge your development with the Google Developer Knowledge API ecosystem“7 Oct 2026, 03:00
- Published by source
- 7 Oct 2026, 03:00
- Found by our system
- 7 Oct 2026, 22:36
- Summary generated
- 7 Oct 2026, 22:38
This article was written by AI from the original source. Facts, numbers and prices come from the source; missing values are marked “Not specified”. Legal notice, copyright and privacy