Use when the operator wants their next prompts routed through Gemini before the agent acts — "enable gemini proxy", "route my prompts through Gemini", "Gemini rewrites my prompts", "включи гемини-прокладку", "переделывай мои промты через гемини", or a /gemini-proxy call; also when they ask to turn the protocol on for the next N prompts or to tune which prompts it covers. Builds on gemini-pro; distinct from one-off gemini-pro / gemini-review / gemini-research.
Use when delegating an arbitrary reasoning or text task to Google Gemini that isn't a structured review or research job — "use Gemini", "ask Gemini Pro", "юзай Gemini", a one-off question, a freeform rewrite or explanation, or offloading heavy reasoning to save the main model's credits/context. For code/plan review use gemini-review; for digesting a large corpus use gemini-research. Runs on the user's Google AI Pro sub via the Antigravity CLI (agy). Text/reasoning only.
Use when the user wants Google Gemini to research, digest, or synthesise over a large body of context — summarise a big file or whole module, compare options/libraries, extract findings across many files, or produce a structured report. Triggers: "ask Gemini to research X", "let Gemini dig into this", "юзай Gemini для ресёрча". Runs on the user's AI Pro sub via the Antigravity CLI (agy). CLI reasoning over context you provide — not the web "Deep Research" agent.
Use when the user wants an independent second-opinion or adversarial review of code, a diff, a pull request, an architecture decision, or a plan from Google Gemini — e.g. "let Gemini review this", "what does Gemini think of this approach", "юзай Gemini для ревью", or when you want model diversity to catch bugs/risks one model alone would miss before shipping. Runs on the user's AI Pro subscription via the Antigravity CLI (agy).