PublicMITGrok-only
X Prompt Gateway × SlimeTree-RLM
▸ Operate X safely via your in-house AI
Sign in to X → send a prompt → the in-browser RLM issues a D/µ/R verdict → only R is delegated to Grok (xAI). Under X's access policy, Grok-only is the optimal choice for processing X data.
1. Sign in to X
Start with an X User Access Token (issue one at developer.x.com) or pick "Demo Guest". Tokens stay in localStorage and never transit javatel.
DEMO
―
2. Grok (xAI) API key ― the sole LLM delegation target
Platform-native LLM principle: processing data on X is exclusive to Grok (xAI native). Avoids X API access restrictions that apply when using other LLMs.
2.5 Operating mode ★ NEW: Pattern B (within-Grok tier)
3. Send a prompt (D=instant / µ=suppressed / R=Grok)
D: 0
µ: 0
R: 0
Grok calls: 0
cheap (B): 0
escalated (B): 0
esc rate (B): ―
Type a prompt and the RLM verdict + response will appear here.
4. Why Grok-only
- X access-policy aligned: avoids restrictions that apply when processing X data with competitor LLMs (Gemini/Claude/OpenAI)
- D instant: 0-token immediate reply (FAQ etc.)
- µ suppressed: speculative / sensitive prompts mechanically blocked
- R Grok: only the prompts that genuinely need an LLM go to xAI
- WAL chain: SHA-256 audit, regulator-ready