grok_realtime
Skill897 starsChanged 4 months agoMITArchived repositoryView on GitHub
xAI Grok Voice provider for Live Conversation.
- Reads credentials
2 other projects use a version of this.
Adopted by
Steady: 0 new projects took it in the last 4 weeks, none the 4 weeks before.thin data
Most recently taken up by
- razzant/ouroborosword for word, first seen Nov 8, 2026
Other versions2 projects use a version of this.
Word for word · 2
- razzant/ouroboros★ 1.7k100% same
- joi-lab/ouroboros-desktop★ 209100% same
Edited · 0
None found.
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository11 other files it gives its agents.
- docs/llms.txtllms.txt
- Browser speechSkill
- Device voiceSkill
- gemini_liveSkill
- live_conversationSkill
- openai_realtimeSkill
- telegramSkill
- TemplatesSkill
- ui_designSkill
- unix_computer_useSkill
- windows_computer_useSkill
The file
Files
Registers grok_realtime:grok for ordinary host-selected context and tools.
Grok Voice 3.0 declares a conversation window of
80,000 tokens in Ouroboros's own estimate (characters / 4), based on operator
probes on 2026-10-21. The 231,828 o200k-token accepted input is separate evidence,
not a fit target. This package never selects or cuts memory. Host facts keep their roles;
Respond opens the greeting only after setup acknowledgements. Call instructions
follow host system text; language tags add a recognition sentence, not a session
field. Silent tool outputs do not request another response.
Selected history containing tool calls is carried as ordered, quoted JSON in
instructions, preserving arguments and results without reissuing those calls.
The provider reports modality usage when present, final response status, and
cumulative connected seconds. Idle samples and the closing tail reach the host
through the observe("usage", ...) callback; the host prices only growth within
one physical connection. Provider usage remains raw evidence beside those axes.
Server VAD decides pauses and interruptions. There is no image input; Show is
unavailable. The feature owns playback, UI and call renewal; the host owns tools,
history, shared spending and Stop. Credentials are read only on an admitted
connection. Installing or granting this provider does not enable Calls (beta).
Catalog checked 2026-10-21: thinking follows reasoning.effort (none/high);
eve is documented and xai_ara was observed as the server default on 2026-10-20.
The empty voice choice leaves the provider default in effect.
This provider declares silent context updates: fitted catch-up history with
respond=False does not open an answer, allowing the host to prepare a successor.
Host-only bookkeeping is absent from the structured wire projection; content strings stay exact. Delivered setup and update rows become exposure evidence only at a usable response, before queued updates are released. Host fitting owns the shared remedies and exact physical-cut checkpoints; the provider does not choose a memory horizon.
A metered read preserves cancellation when its native result arrives at teardown; the connection reader always joins its pending read.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

