📄 SKILL.md

← 返回目录


name: qianwen-ops-auth description: "[QianWen] Configure authentication (API keys, endpoints). TRIGGER when: setting up QIANWEN_API_KEY, troubleshooting 401/auth errors, when another skill reports missing credentials, or user explicitly invokes this skill by name (e.g. use qianwen-ops-auth). DO NOT TRIGGER when: non-auth Qwen tasks, general API usage questions." compatibility: "Requires curl for verification. Cursor: auto-loaded. Claude Code: read this skill's SKILL.md before first use."

Agent setup: If your agent doesn't auto-load skills (e.g. Claude Code), see agent-compatibility.md once per session.

QianWen Authentication Setup

Configure and verify authentication for QianWen APIs. This skill is part of QianWen-AI/qianwen-ai.

Skill directory

Use this skill's internal files for learning. Load references only when the user needs console or documentation links.

| Location | Purpose | |----------|---------| | references/tokenplan.md | Token Plan 团队版 vs standard key: endpoint mapping, supported models (4 text + 4 image), Credits billing, error codes | | references/custom-oss.md | Custom OSS bucket setup for production file uploads (replaces 48h temp storage) | | references/sources.md | Console URLs, auth guide (manual lookup only) | | references/agent-compatibility.md | Agent self-check: register skills in project config for agents that don't auto-load |

Security

NEVER output any API key, OSS credential in plaintext. This applies equally to DASHSCOPE_API_KEY and custom OSS AccessKey pairs. Any check or detection of credentials in this skill must be non-plaintext: report only status (e.g. "set" / "not set", "valid" / "invalid", HTTP status code), never the key value.

API Key Handling (MANDATORY)

When the API key is not configured or a script reports missing credentials:

1. NEVER ask the user to provide their API key directly. Do not prompt "please paste your API key" or similar. Do not request the key value in any form. 2. Help create a .env file with a placeholder, then instruct the user to fill in their own key: - Run: echo 'DASHSCOPE_API_KEY=sk-your-key-here' >> .env - Tell the user: "Please replace sk-your-key-here with your actual API key from the QianWen Console." 3. Or explain how to configure the environment variable: export DASHSCOPE_API_KEY='sk-...' + provide the console URL. 4. Only write the actual key value into .env if the user explicitly insists on having the agent do it for them.

Credential Priority Chain

Credentials are loaded in the following order (first match wins):

1. Environment variableDASHSCOPE_API_KEY (or QIANWEN_API_KEY alias) 2. .env file — in current working directory, then repo root (detected via .git or skills/ directory). Existing environment variables are not overwritten.

Environment Variables

| Variable | Purpose | |---------------------|-------------------------------------------------------------------------------------------------------------------------------------------| | DASHSCOPE_API_KEY | API key (required) | | QIANWEN_API_KEY | Alias for DASHSCOPE_API_KEY. If both are set, QIANWEN_API_KEY takes priority. | | QWEN_BASE_URL | Override default endpoint (optional; for custom deployments) | | QWEN_TMP_OSS_BUCKET | Custom OSS bucket for file uploads (replaces 48h temp storage). See custom-oss.md. | | QWEN_TMP_OSS_REGION | OSS region (required when QWEN_TMP_OSS_BUCKET is set). | | QWEN_TMP_OSS_AK_ID / AK_SECRET | OSS credentials (use RAM user with least-privilege: oss:PutObject + oss:GetObject). Falls back to OSS_ACCESS_KEY_ID / OSS_ACCESS_KEY_SECRET if not set. |

API Key Types

QianWen has two mutually exclusive key types:

| Key Type | Format | Purpose | Endpoint | |----------|--------|---------|----------| | Standard (Pay-as-you-go) | sk-xxxxx | API calls from scripts, apps, and tools | dashscope.aliyuncs.com | | Token Plan (团队版) | sk-sp-xxxxx | Interactive AI tools only (Cursor, Claude Code, Qwen Code, OpenClaw, OpenCode, Codex, Kilo Code/CLI, Hermes Agent) | token-plan.cn-beijing.maas.aliyuncs.com |

All QianWen-AI/qianwen-ai execution scripts require a standard sk- key on dashscope.aliyuncs.com. Token Plan keys (sk-sp-) cannot be used by these scripts — they target a different endpoint (token-plan.cn-beijing.maas.aliyuncs.com) and produce InvalidApiKey: Invalid API-key provided on standard endpoints. Token Plan 团队版 supports only 4 text LLMs (qwen3.6-plus, glm-5, MiniMax-M2.5, deepseek-v3.2) plus 4 image generation models (qwen-image-2.0, qwen-image-2.0-pro, wan2.7-image, wan2.7-image-pro, accessed via tool Skill/extension mechanism only). Video, TTS, ASR, embedding, and other modalities are not supported.

If the user's key starts with sk-sp-, guide them to obtain a standard key from the console below. See tokenplan.md for full details.

Viewing Bills

Use the qianwen-usage skill to query usage, free tier quota, and billing directly. Alternatively, billing details are available in the QianWen console:

| Key Type | Billing Page | |----------|--------------| | Standard (Pay-as-you-go) | Pay-as-you-go Billing | | Token Plan 团队版 | Token Plan Subscription | | Usage analytics (Pay-as-you-go) | Usage Analytics |

NEVER fabricate, guess, or construct usage/billing/console URLs. Only provide the exact links listed in this skill. If a URL is not listed here, do not invent one.

Getting an API Key

1. Open the QianWen Console 2. Sign in with your QianWen account 3. Create or copy an API key from the API Key management section 4. Standard keys start with sk- (not sk-sp- which is Token Plan 团队版 only)

Security Best Practices

- Never hardcode API keys in source code or config files committed to version control

The script automatically loads .env from the current working directory and the project root (detected via .git or skills/ directory). Existing environment variables are not overwritten by .env values.

Example .gitignore entry

.env
.env.local
*.env

Verification

Unless explicitly stated otherwise, any script or task mentioned in this skill runs in the foreground — wait for standard output; do not run it as a background task.

Test authentication with a simple curl request:

curl -sS -X POST "https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions" \
  -H "Authorization: Bearer $DASHSCOPE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"qwen-turbo","messages":[{"role":"user","content":"Hi"}]}'

A successful response returns JSON with choices and message.content.

Authentication Error Handling

QianWen API keys are scoped to the QianWen console. An invalid or mismatched key produces 401 Unauthorized.

When to trigger

When any sub-skill receives a 401 response and a non-plaintext check shows the key is set (e.g. [ -n "$DASHSCOPE_API_KEY" ]; do not output the key value).

Probe command

Send a lightweight request to verify authentication:

curl -sS -o /dev/null -w "%{http_code}" \
  -X POST "https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions" \
  -H "Authorization: Bearer $DASHSCOPE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"qwen-turbo","messages":[{"role":"user","content":"hi"}]}'

On 401: mandatory interactive resolution

If the probe returns 401, follow these steps in order:

Step 1 — Confirm the key origin:

Your API key failed authentication.

Please confirm: 1. Your key was created at platform.qianwenai.com/home (QianWen console) → re-verify the key 2. My key may be invalid → create a new one at platform.qianwenai.com/home/api-keys

Step 2 — Apply the user's selection:

| User says | Action | |-----------------------------------|---------------------------------------------------------------------| | Key is from QianWen console | Re-run verification to confirm the key works | | Key may be invalid | Guide user to create a new key at platform.qianwenai.com/home/api-keys |

Step 3 — Re-run verification.

Step 4 — If still 401, the key itself is invalid — guide the user to create a new one from the QianWen console.

Validation

Run the curl verification command above. Pass criteria: HTTP 200 response with valid JSON containing choices and message.content. Save output to output/qianwen-ops-auth/ for evidence.

Output And Evidence

- Save verification results and any diagnostic output to output/qianwen-ops-auth/.

2. If found — run: python3 /scripts/check_update.py --print-response and report the result. Use --force if the user asks to force-check. 3. If not found — you MUST present the following options and wait for the user's response:

The update-check skill is not installed. It keeps you informed about new QianWen-AI/qianwen-ai features and bug fixes.
>
1) Install now — I'll run the install command for you
2) Skip for now — do nothing this time
3) Never remind me — permanently suppress this prompt

After the user responds: