All Articles
The Image Model Shuts Down on October 29 — Find Every Call in Your Agent Setup and Cut Over with a Way Back
The Gemini API deprecations table lists a shutdown date of October 29 for the image model gemini-3.1-flash-image. Here is how I locate every call scattered across configs and code, match them to a ledger, and fix a cutover date and a fallback in advance — with scripts that run.
antigravity-preview-05-2026 shuts down October 5: which setups need more than a string swap
The shutdown date for antigravity-preview-05-2026 is set for October 5, and the official page gives you the date and the successor name, nothing more. If you only read output_text, swapping the string is the whole migration. If you read steps, it is not.
When a session starts with MCP tools missing, where do you stop?
When MCP initialization fails at startup, the agent quietly falls back to other means and keeps going. Here is a small script that asks your servers directly which tools they expose, plus the three branches I decided on in advance for when something is missing.
Antigravity CLI 1.1.28 turns a print-timeout overrun into a successful exit — rechecking unattended runs
Since 1.1.28, an expired --print-timeout returns partial output and exits successfully instead of failing. Fetching an external URL now asks for approval first. Here is how I rebuilt an unattended pipeline that judged everything by the exit code.
Pointing GOOGLE_GEMINI_BASE_URL at Your Own Gateway Does Not Change Who Holds the Key
I pointed the Antigravity CLI at a gateway of my own and recorded exactly which credentials travelled with it. Authorization is dropped across a cross-origin redirect, but the API key header survives every hop. Here is the measurement, and how I moved the key off my laptop.
Adding MCP Servers Without Opening the Config File: The agy mcp Subcommand
Antigravity CLI 1.1.16, released on August 20, adds an mcp subcommand so you can register and toggle MCP servers without hand-editing JSON. Here is how to add stdio and HTTP servers, isolate a misbehaving one, and script your whole setup so it travels to a new machine.
Driving Antigravity CLI with stream-json Input: Where to Split the Conversation
The print mode now accepts --input-format stream-json, letting an external driver hold one session open and feed it jobs one at a time. I measured how much a single conversation carries across 24 jobs, found that a byte budget splits work in the wrong place, and worked out how to choose the split boundary instead.
Find the one broken MCP entry before Antigravity starts, with 40 lines of Node
A single typo in your MCP settings can take every server down with it. Here is a small preflight script that catches the mistakes statically, the real output from a deliberately broken config, and the duplicate-key trap that JSON hides from you.
The MCP Server I Thought I Killed Was Still Holding the Port: 84 Teardown Trials
Closing a session should stop the MCP servers it started. Across 84 trials of four teardown strategies, a process-group SIGTERM left zero orphans — until the child called setsid, at which point it left 100%. Includes a full descendant-sweep implementation.
Startup Dropped to 3ms, and the First Turn Saw Zero Tools
Non-blocking MCP loading cut startup from 2,907ms to 3ms. The wait did not disappear — it moved into the first turn. Here is where it went, measured, plus a readiness gate that waits only for what a turn actually needs.
I Measured Before Writing a Number: MCP Connect and Tool Calls Differed by 486x
Antigravity 2.4.3 lets you set a timeout per MCP server. To find a defensible number I built a stdio server, measured each boundary separately, and found why a single value cannot cover both.
Three Tools Named read_file: Catching MCP Name Collisions Before Startup
Bundle enough MCP servers and tool names collide quietly. Here is what a real 5-server, 21-tool setup measured at 43% collision, plus a Python preflight that catches them before startup and assigns deterministic aliases.