Stats
Actions
Tags
Help us improve
Share bugs, ideas, or general feedback.
From multi-model
Opus auto-routes every task to the right model without asking. Triggers on ANY non-trivial request — planning, coding, refactor, review, research, audit, debug, multi-file work. Opus plans + synthesizes only; Sonnet/Haiku/Ollama/NVIDIA/Codex execute in parallel; Codex verifies.
npx claudepluginhub ranjankumarpatel/claude-code-multi-model --plugin multi-modelHow this skill is triggered — by the user, by Claude, or both
Slash command
/multi-model:orchestratorThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
**Rule**: Opus decides the model per task from the request signal. Never ask the user which model to use. Never default to "ask first". Route silently, report briefly.
Weighted social-graph ranking for warm intro discovery, bridge scoring, and network gap analysis across X and LinkedIn. Use when the user wants the reusable graph-ranking engine itself, not the broader outreach or network-maintenance workflow layered on top of it.
Share bugs, ideas, or general feedback.
Rule: Opus decides the model per task from the request signal. Never ask the user which model to use. Never default to "ask first". Route silently, report briefly.
Agent tool.mcp__ollama__ollama_chat): gemma4:31b-cloud, kimi-k2.5:cloud, kimi-k2-thinking:cloud.mcp__nvidia-nim__nvidia_chat): nemotron-ultra, nemotron-super, deepseek-r1, llama405b, mistral-large, gemma4.mcp__nvidia-security__nvidia_security_chat): audits, PII, guardrails./codex:review, /codex:adversarial-review, codex:codex-rescue subagent.| Task signal | Route to |
|---|---|
| Bulk read, grep, rename, format, list files | Haiku |
| Multi-file refactor, template logic, debugging, test writing | Sonnet |
| Deep reasoning / chain-of-thought / hard logic puzzle | kimi-k2-thinking:cloud or deepseek-r1 |
| Coding w/ second opinion or alt-frontier | gemma4:31b-cloud or nemotron-ultra |
| Long-context / agentic / vision+language | kimi-k2.5:cloud |
| Multilingual / non-English code | mistral-large |
| Large general-purpose | llama405b |
| Security audit, CVE, OWASP, SAST, PII, prompt-injection, compliance | NVIDIA Security |
| Stuck / failing tests / adversarial review / pre-merge verify | Codex |
| Independent subtasks (≥2) | Parallel Agent calls in ONE message |
/codex:review on the diff before declaring done.State route chosen + why in one line per dispatch. Example:
Routing: refactor → Sonnet; bulk rename → Haiku; security scan → NVIDIA Security (parallel).