This Cross-AI page is English only. Language controls on other site pages are unchanged.
Experimental building blocks. Note: Connecting each AI may require adjustments on your side (API keys, model choice, prompts, and runtime settings). Treat these desks as experimental building blocks, not a guaranteed turnkey multi-cloud product.
Cross-AI
Cross-AI workflow foundation
Adaptable workflow foundation — sequential envelopes, conflict logs, and Human Yes hard-stops — not a finished multi-cloud service. Live experimental desks: Research Desk, Operations Desk, and Software Team (PARTIAL). Content/Trading stay HOLD.
Cross-AI vs Marketplace vs Library
Cross-AI
Adaptable orchestration pattern for honest seats. An LLM model and process tools (fetch, file audit, coordinator, stop-gate) are labeled separately. Sequential envelopes, conflict logs, Human Yes gates. Public proof — not a Library Verified badge and not a multi-provider marketing claim until those providers actually run. Only Research ships a distributable local kit; Ops/Software are case-study + proof surfaces you adapt yourself.
Marketplace
Ready-to-use AI teams / paid desks buyers run or buy (Teams, AI team marketplace). Separate commercial surface. Cross-AI does not sell itself as a Marketplace listing.
Library (Build Library)
Free local recipes (Ollama Modelfiles, workflows). Untested until an operator marks tested. Cross-AI may use Library models as seats; it does not flip Verified.
Live now
LIVE Research Desk, Operations Desk, and Software Team (PARTIAL) are published.
Research Desk
Answer quality: 0 / 9 accepted (September 11 battery; incomplete transport counted separately). Kit workflow can run; answers are not reliable for unattended research, publication, or trading.
Experimental local workflow — answer-quality evaluation failed. The September 11 review found unsupported publication approvals, false claim attributions and missed numerical inconsistencies. The earlier PASS refers to limited workflow checks, not answer reliability. This kit is for controlled experiments, not unattended research, publication approval or trading decisions. Outputs require checking against the original sources. No external action was executed in these tests. This is sequential local Ollama use, not simultaneous collaboration between cloud AI services.
One Ollama LLM plus process seats in historical workflow checks. Local run kit available.
Operations Desk
Switchboard-style triage with stop-gate. External reply STOPPED at Human Yes and was NOT sent. RESULT PASS · SAMPLE ticket.
Software Team
PARTIAL / UNVERIFIED. Local demo + Ollama coding-review ran. Claude Code / CloudAgent UNAVAILABLE — not invented. No push/merge/deploy.
HOLD — not published
These packages are ranked HOLD and have no product pages live:
- Content Desk — design stub
- Trading Research Desk — design stub · ANALYSIS≠EXECUTION
Support-triage PE elevate remains HOLD. GrokHub ingest is NO-GO. No Verified flip from Cross-AI.
What this is not
- Not a Library Verified claim
- Not simultaneous multi-AI debate theater
- Not a send / mass-DM / X ops surface
Research Desk Operations Desk Software Team Library Marketplace Teams