No user-submitted related servers found.
Scored across 17 tools
The three main Gemini tools are explicitly differentiated (gemini_ask = read-only local reasoning, gemini_research = web with citations, gemini_code = edits a local folder), and the Jules lifecycle tools each map to one step (start/status/message/approve_plan/sessions/sources). The main residual overlap is that gemini_ask is also described as able to search the web, blurring it slightly against gemini_research, and video_generate references browser-only tools (flow_make_video, flow_collect) that are not in this set, which can send an agent looking for tools that aren't there.
Names are all snake_case and organized into clear namespaces (gemini_*, jules_*, video_*, image_generate, google_pro_benefits), so grouping is predictable. The only wrinkle is the mixed word order: gemini_ask/gemini_code/jules_start are prefix_action while image_generate/video_stitch are action-noun, plus video_check is not an action verb—minor inconsistency but still readable.
Seventeen tools is on the heavy side but the server spans several genuinely different capabilities (chat/research, local code editing, image gen, video gen/edit/stitch, Jules cloud sessions, plus model/agent discovery), and most tools reflect a real lifecycle step. A few, like google_pro_benefits and the separate video_check, could arguably be folded in, but none feels gratuitous.
Async workflows are covered end to end: video_generate/video_edit return ids, video_check fetches them, video_stitch joins clips, and Jules has sessions/sources → start → status → message → approve_plan. Gaps are minor and workaroundable: no tool to list or clean up artifacts already in GEMINI_PRO_OUTPUT_DIR, no image_check (image generation is synchronous, so acceptable), and the advertised Flow/browser tools are absent unless the server is started with GEMINI_PRO_ENABLE_BROWSER=1.