“Philipp Schmid at Google I/O demoed how to build an AI agent with its own secure, hosted Linux sandbox in a single API call using Gemini Managed Agents and the new Interactions API to execute code and manage its memory.”

#8 𝕏 Philipp Schmid at Google I/O demoed how to build an AI agent with its own secure, hosted Linux sandbox in a single API call using Gemini Managed Agents and the new Interactions API to execute code and manage its memory. #9 𝕏 Philipp Schmid clarifies that each sandbox session gets its own persistent container ID—shareable across agents—and that they’re currently trading persistence for slower cold starts.

“Philipp Schmid built a GitHub Issue Triage Agent using a single curl to the Gemini API.”

#9 𝕏 Philipp Schmid built a GitHub Issue Triage Agent using a single curl to the Gemini API.

“Philipp Schmid published a new guide for streaming in the Gemini Interactions API to make building streaming applications super easy.”

#12 𝕏 Philipp Schmid published a new guide for streaming in the Gemini Interactions API to make building streaming applications super easy. Just point your agent to it and let it handle the rest.

“#2 𝕏 Philipp Schmid announced that Android 16+ now includes a built-in on-device MCP, letting developers tag app functions with a new `@AppFunction` annotation so agents like Gemini can call and chain them across apps without any server or network calls.”

#2 𝕏 Philipp Schmid announced that Android 16+ now includes a built-in on-device MCP, letting developers tag app functions with a new `@AppFunction` annotation so agents like Gemini can call and chain them across apps without any server or network calls. #3 📝 OpenAI News Building a safe, effective sandbox to enable Codex on Windows - OpenAI built a custom, unelevated sandbox for Codex on Windows after finding AppContainer too narrowly scoped, Windows Sandbox incompatible with acting on the user's real checkout (and unavailable on Home), and MIC unsafe because relabeling workspaces lowers their integrity; the prototype gives the sandbox a distinct identity via synthetic SIDs and uses write-restricted tokens to limit where Codex can modify files. The default Codex sandbox runs commands with reduced permissions—allowing reads broadly, writes only inside the workspace, and no internet access unless explicitly enabled.

“#7 𝕏 Philipp Schmid published a guide for Gemini Interactions API’s new `thought` steps and encrypted signatures, detailing stateful vs. stateless modes, seamless model switching, and effortless context management to supercharge agent development.”

#7 𝕏 Philipp Schmid published a guide for Gemini Interactions API’s new `thought` steps and encrypted signatures, detailing stateful vs. stateless modes, seamless model switching, and effortless context management to supercharge agent development. Also covered by: @Sundar Pichai

“Philipp Schmid shares Google’s Gemini API interactions quickstart guide, helping PM builders quickly set up and test the new Gemini AI model.”

#20 𝕏 Philipp Schmid shares Google’s Gemini API interactions quickstart guide, helping PM builders quickly set up and test the new Gemini AI model. #21 𝕏 Lenny Rachitsky shares eight actionable insights from Eric Ries—spanning financial gravity, CEO retention post-IPO, public-benefit corp structures like AnthropicAI, mission protection, and principled decision-making exemplified by Cloudflare.

“#4 𝕏 Philipp Schmid updated the Gemini Interactions API to replace rigid `user`/`model` roles with discrete “steps” (user_input, thought, function_call, tool_call, model_output, etc.), consolidated response_format controls, and added a toggle in the docs.”

The item describes API changes and new interaction structure for Gemini.

“Philipp Schmid : The Gemini API File Search tool now offers true multimodal PDF and image retrieval using `gemini-embedding-2`, handling chunking, embedding, indexing and grounding in one call.”

#4 𝕏 Philipp Schmid : The Gemini API File Search tool now offers true multimodal PDF and image retrieval using `gemini-embedding-2`, handling chunking, embedding, indexing and grounding in one call. #15 𝕏 Philipp Schmid shows Gemma 4 pushing the Pareto frontier on Code Arena, with Gemma-4-31b at #13 and Gemma-4-26b-a4b at #17 among open models you can run on a MacBook Pro.

“Philipp Schmid launched Multi-Token Prediction for Gemma 4, tripling inference speed with zero quality loss—now available E2B/E4B under Apache 2.0.”

#13 𝕏 Philipp Schmid launched Multi-Token Prediction for Gemma 4, tripling inference speed with zero quality loss—now available E2B/E4B under Apache 2.0. #14 𝕏 Philipp Schmid outlines four subagent coordination patterns—tool calls, spawns, pools, and teams—to structure multi-agent workflows.

“#9 𝕏 Philipp Schmid launched a QoL upgrade for the Interactions API: errors now name the exact field and bad value, list supported enum options, show expected vs. actual formats, and pinpoint field paths like `input[0].name`.”

#8 𝕏 Logan Kilpatrick rolled out extensive error message improvements for the Interactions API, making its feedback far more human- and agent-readable. #9 𝕏 Philipp Schmid launched a QoL upgrade for the Interactions API: errors now name the exact field and bad value, list supported enum options, show expected vs. actual formats, and pinpoint field paths like `input[0].name`. #10 𝕏 Guillermo Rauch launched npx deepspec, an open-source agent orchestrator that leverages thousands of parallel coding agents in Vercel Sandbox to uncover critical security vulnerabilities in minutes. #11 ▶️ AI Agents run my business and life Greg Isenberg Andrew Wilkinson demonstrates running an autonomous SaaS business using Claude-based OpenClaw agents orchestrated in Harbor to handle support ticket triage (including auto-fixing P0 issues and merging PRs) and marketing campaigns via Post Hog and Meta/Reddit ads.