The context · April 2025

Google introduced the Agent2Agent protocol in April 2025 to support communication and collaboration between agents built with different tools and frameworks. [1]

A proposed project brief inspired by the period’s developments. This is an idea for exploration, not an announced Sansa product, an approved roadmap or a claim of completed client work.

Keep the task modest

Imagine a service assistant asking a documentation assistant to find an approved technical answer. The first assistant supplies a sanitised question; the second returns an answer with references and an explicit unresolved status when the material is insufficient. Neither changes a customer record.

Test the awkward exchanges

Exercise missing references, expired requests, duplicate requests and a receiving service that is unavailable. Give each exchange a traceable identifier. Check that a repeated message does not create duplicate work and that an incomplete response reaches a person rather than being presented as settled fact.

A decision-ready outcome

This concept would produce a working handover prototype, example traces and a short ownership agreement. Success means reviewers can follow where an answer came from and know what to do when it fails. Additional agents should be introduced only when they remove a concrete constraint in that workflow.

Source & context

Google Developers · Agent2Agent, 9 April 2025

This retrospective was written for the archive in September 2026. The linked primary source documents the announcement or event; the practical interpretation and proposed approach are Sansa’s editorial perspective. Public examples do not imply a client relationship. Product capabilities and guidance may have changed since the period discussed.

Another perspective · April 2025

When agents can talk, who owns the handover?

Continue reading

Working through a similar question?

Talk it through with Sansa