qwencloud-model-selector
[QwenCloud] Recommend the best Qwen model and parameters. TRIGGER when: choosing between Qwen models, comparing Qwen model pricing, understanding Qwen model capabilities, checking usage or billing, viewing cost history, when an execution skill needs model selection advice, or user explicitly invokes this skill by name (e.g. use qwencloud-model-selector). DO NOT TRIGGER when: non-Qwen model discussions (OpenAI, Gemini, etc.), general AI questions unrelated to Qwen.
What this skill does
> **Agent setup**: If your agent doesn't auto-load skills (e.g. Claude Code),
> see [agent-compatibility.md](references/agent-compatibility.md) once per session.
# Qwen Model Selector (Advisor)
This skill operates in two modes:
1. **Interactive advisory** — asks diagnostic questions to recommend the right model (see Diagnostic Flow).
2. **Cross-skill resolution** — provides a fast-path model lookup for execution skills that need a model
decision without user interaction (see [recommendation-matrix.md](references/recommendation-matrix.md)).
Do not fabricate model names — only recommend models listed in this skill or returned by CLI.
This skill is part of **qwencloud/qwencloud-ai**.
## Skill directory
Load on demand. Do not fetch external URLs unless the user explicitly asks for the latest data.
| Location | Purpose |
|-------------------------------------------|----------------------------------------------------------------------------------|
| `references/cli-usage.md` | **CLI-first data strategy**: when to use CLI, 3-step login flow, display rules |
| `references/error-handling.md` | CLI error classification & recovery actions (auth, not-found, network, ...) |
| `references/recommendation-matrix.md` | Full model recommendation tables, Cross-Skill Resolution, Coding Plan, Thinking |
| `references/pricing-disclaimer.md` | Pricing guidance + **mandatory** cost-estimation disclaimer (CN/EN) + console links |
| `references/pricing.md` | Pricing structural overview (offline snapshot) |
| `references/model-list.md` | Model catalog (offline snapshot) |
| `references/sources.md` | Official documentation URLs (manual lookup only) |
| `references/agent-compatibility.md` | Agent self-check for skill registration |
## Prerequisites
**QwenCloud CLI is strongly recommended** — it is the authoritative real-time data source for model
availability, pricing, and quotas. Verify with:
```bash
qwencloud version
```
If not installed:
```bash
npm install -g @qwencloud/qwencloud-cli
```
Node.js >= 18 required. Without CLI you can still answer general navigation questions from offline
snapshots, but **you cannot answer "latest", "exact price", or "specific model details" questions**.
## Security & Credential Model
QwenCloud has **two independent credential systems** — never confuse them:
| Credential | Purpose | How to provide |
|------------|---------|----------------|
| **API Key** (`sk-...` / `sk-sp-...`) | Call model APIs in your code | `$DASHSCOPE_API_KEY` / `$QWEN_API_KEY` env var |
| **CLI session** | Authorize `qwencloud` CLI subcommands | `qwencloud auth login` (browser device flow) |
**Red lines (apply to both):**
- **NEVER output any credential value in plaintext.** Use variable references; report only status
("set" / "not set", "valid" / "invalid"). Never display `.env` or config file contents.
- **NEVER conflate the two systems.** When CLI returns `Not authenticated` / `AUTH_REQUIRED`, run the
3-step device-flow login (see [cli-usage.md](references/cli-usage.md#authentication-3-step-login-flow)).
**DO NOT** ask the user for an API key, and **DO NOT** try to set `$DASHSCOPE_API_KEY` to fix CLI auth.
## Data Resolution Order
Match the user's question to the right data source. **Do not fall back to a lower tier without trying
the recovery actions in the higher tier first.**
| Question type | Primary source | Notes |
|----------------------------------------------------------------|---------------------------------------------------------|------------------------------------------------------|
| General navigation ("which family for text chat?") | SKILL.md `Default` table + `recommendation-matrix.md` | Offline-answerable |
| **Latest / exact / specific** (price, model details, quota) | **CLI MUST be used** — see `cli-usage.md` | Snapshots are stale; never invent numbers |
| Search by capability ("model that does X") | `qwencloud models search "<X>" --format json` | Snapshot keyword coverage is incomplete |
| CLI returned an error | `error-handling.md` recovery actions, **then retry** | Auth failure → run 3-step login, do not skip to snapshot |
| CLI completely unavailable AND user declines install/login | `model-list.md`, `pricing.md` (with stale-data caveat) | Only after CLI recovery genuinely failed |
| All of the above cannot answer AND user confirms online lookup | URLs in `sources.md` | Never proactively fetch |
## Diagnostic Flow (Interactive Advisory)
Ask the user (in order):
1. **Content type?** — text / image / video / audio / vision
2. **Primary task?** — generation / understanding / coding / reasoning / translation
3. **Priority?** — quality vs speed vs cost
4. **Input size?** — short / medium / long context
5. **Structured output?** — JSON / function calling needed?
## Default Recommendations
No clear signals → use the canonical default for the domain. For specialized cases (reasoning, coding,
OCR, role-play, image editing, etc.) and per-domain comparison, see
[recommendation-matrix.md](references/recommendation-matrix.md).
| Domain | Default | Quality | Speed | Cost |
|---------------------|------------------|------------------|--------------------|--------------------|
| text.chat | qwen3.6-plus | qwen3-max | qwen3.5-flash | qwen-turbo |
| vision.analyze | qwen3.6-plus | qwen3-vl-plus | qwen3-vl-flash | qwen3-vl-flash |
| omni (voice+vision) | qwen3-omni-flash | qwen3-omni-flash | qwen3-omni-flash | — |
| image.generate | wan2.6-t2i | wan2.6-t2i | wan2.2-t2i-flash | wan2.2-t2i-flash |
| image.edit | wan2.6-image | wan2.6-image | wan2.5-i2i-preview | wan2.5-i2i-preview |
| video.t2v | wan2.6-t2v | wan2.6-t2v | — | — |
| video.i2v | wan2.6-i2v-flash | wan2.6-i2v | wan2.6-i2v-flash | — |
| audio.tts | qwen3-tts-flash | cosyvoice-v3-plus| qwen3-tts-flash | qwen3-tts-flash |
> **Degradation**: If this skill is not loaded, each execution skill falls back to its own built-in
> default. This protocol is purely additive — it enhances model selection but never blocks execution.
## CLI Quick Reference
> **Auth required.** All `models` and `usage` commands need an active **CLI session** (browser
> device-flow login — **NOT** the API key). If the command returns `Not authenticated` / `AUTH_REQUIRED`:
> 1. **Run the 3-step device-flow login** in [cli-usage.md](references/cli-usage.md#authentication-3-step-login-flow)
> (proactively open the verification URL using the OS-appropriate command, then poll immediately).
> 2. **Retry the original command** after `success`.
> 3. **DO NOT** ask the user for `$DASHSCOPE_API_KEY` / `$QWEN_API_KEY` — those are for model API
> calls, not CLI session. See [Security & Credential Model](#security--credential-model) above.
> 4. **DO NOT** silently fall back to snapshots.
| Need | Command |
|-------------------------------|--------------------------------------------------------------------|
| Full mRelated in AI Agents
skill-development
IncludedComprehensive meta-skill for creating, managing, validating, auditing, and distributing Claude Code skills and slash commands (unified in v2.1.3+). Provides skill templates, creation workflows, validation patterns, audit checklists, naming conventions, YAML frontmatter guidance, progressive disclosure examples, and best practices lookup. Use when creating new skills, validating existing skills, auditing skill quality, understanding skill architecture, needing skill templates, learning about YAML frontmatter requirements, progressive disclosure patterns, tool restrictions (allowed-tools), skill composition, skill naming conventions, troubleshooting skill activation issues, creating custom slash commands, configuring command frontmatter, using command arguments ($ARGUMENTS, $1, $2), bash execution in commands, file references in commands, command namespacing, plugin commands, MCP slash commands, Skill tool configuration, or deciding between skills vs slash commands. Delegates to docs-management skill for official documentation.
reprompter
IncludedTransform messy prompts into well-structured, effective prompts — single or multi-agent. Use when: "reprompt", "reprompt this", "clean up this prompt", "structure my prompt", rough text needing XML tags and best practices, "reprompter teams", "repromptception", "run with quality", "smart run", "smart agents", multi-agent tasks, audits, parallel work, anything going to agent teams. Don't use when: simple Q&A, pure chat, immediate execution-only tasks. See "Don't Use When" section for details. Outputs: Structured XML/Markdown prompt, quality score (before/after), optional team brief + per-agent sub-prompts, agent team output files. Success criteria: Single mode quality score ≥ 7/10; Repromptception per-agent prompt quality score 8+/10; all required sections present, actionable and specific.
adaptive-compaction
IncludedAdaptive add-on policy and recovery layer that decides WHEN to compact, prune, snapshot, or fork -- replacing fixed-percent auto-compaction across Claude Code, Codex, and MCP-capable hosts. Trigger on auto-compact timing or damage: "when should I compact", "is it safe to compact now or start a fresh session", "auto-compact fires too early/mid-task", "switching to an unrelated task but the window still has space", "context rot", "answers get worse the longer the session runs", "the agent forgot the plan or my decisions after it summarized", "add a layer on top that manages context without changing the agent", raising autoCompactWindow to give the policy room, or installing/tuning a cross-tool compaction policy or PreCompact hook -- even when "compaction" is never said but the problem is context-window pressure or post-summarization memory loss. Do NOT use to summarize a conversation, build RAG, write a summarization prompt (decides WHEN not HOW), or answer max-context-length trivia.
agent-skill-creator
IncludedCreate cross-platform agent skills from workflow descriptions. Activates when users ask to create an agent, automate a repetitive workflow, create a custom skill, or need advanced agent creation. Triggers on phrases like create agent for, automate workflow, create skill for, every day I have to, daily I need to, turn process into agent, need to automate, create a cross-platform skill, validate this skill, export this skill, migrate this skill. Supports single skills, multi-agent suites, transcript processing, template-based creation, interactive configuration, cross-platform export, and spec validation.
llm-wiki
IncludedUse when building or maintaining a persistent personal knowledge base (second brain) in Obsidian where an LLM incrementally ingests sources, updates entity/concept pages, maintains cross-references, and keeps a synthesis current. Triggers include "second brain", "Obsidian wiki", "personal knowledge management", "ingest this paper/article/book", "build a research wiki", "compound knowledge", "Memex", or whenever the user wants knowledge to accumulate across sessions instead of being re-derived by RAG on every query.
skill-master
IncludedAgent Skills authoring, evaluation, and optimization. Create, edit, validate, benchmark, and improve skills following the agentskills.io specification. Use when designing SKILL.md files, structuring skill folders (references, scripts, assets), ingesting external documentation into skills, running trigger evals, benchmarking skill quality, optimizing descriptions, or performing blind A/B comparisons. Keywords: agentskills.io, SKILL.md, skill authoring, eval, benchmark, trigger optimization.