Skip to main content
skill CX Codex

qwencloud-usage — Claude Code skill: usage, billing, subscriptions

A skill for your QwenCloud account: login and logout, credit usage, billing, quotas, call logs and subscriptions. Useful when a request failed or you need to know what's left.

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add QwenCloud/qwencloud-ai --skill qwencloud-usage -a claude-code
Open on GitHub

What qwencloud-usage does

The skill covers everything tied to a QwenCloud account: authentication, usage, billing, call logs, quotas, subscriptions, orders, team seats and the PAYG spending limit. It won't invent or mock account data — if something isn't there, it says so.

It works in steps. First it reads references/cli.md, the only source for installation, the authentication flow, commands, arguments, pagination and output modes. Then it opens one result-semantics reference for the question at hand: references/usage.md for usage and logs, references/billing.md for billing cycles and the spending limit, references/subscriptions.md for subscriptions and seats, references/authentication.md for login and expired credentials. An exact requestId lookup runs without --from, --to or --period — those go in only when you ask for them.

The output rules are strict. Token Plan Credits keep full precision, no rounding: 63028.84805624 stays as it is. Coding Plan never appears in user-facing output, not even as a "not subscribed" placeholder. If a top-N limit cuts rows or pagination is unfinished, the result is marked partial. Billing and subscription figures always carry the note that final amounts come from the console billing statement. Worth knowing if you want a number to pay: the skill returns API data, not your final invoice.

What it helps with

  • log in to a QwenCloud account and out again
  • check credit usage and the Token Plan balance
  • find which requests failed with 4xx or 5xx
  • look up a single request by requestId
  • review the PAYG spending limit and free tier quotas

Example prompts for the agent

  • show my usage this week and the remaining credits
  • look up this requestId and tell me why it failed
  • how many team seats are free, and when do credits refresh

How to install qwencloud-usage

1Skills CLI (Vercel)

npx skills add QwenCloud/qwencloud-ai --skill qwencloud-usage -a claude-code

2Manually: put the files in place

~/.claude/skills/qwencloud-usage/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/qwencloud-usage/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about qwencloud-usage

How do I install qwencloud-usage in Claude?

Run npx skills add QwenCloud/qwencloud-ai --skill qwencloud-usage -a claude-code. The skill lands in ~/.claude/skills/qwencloud-usage/SKILL.md.

What does qwencloud-usage do?

It handles QwenCloud account authentication and answers questions about usage, billing, call logs, quotas, subscriptions, orders and team seats. Before answering, it reads the references under references/.

How is qwencloud-usage different from qwencloud-payment?

This skill doesn't touch payment methods or model browsing — that's qwencloud-payment. Here you get login, usage, billing, logs and subscriptions.

Source text

SKILL.md · Apache-2.0
---
name: qwencloud-usage
description: "Manage account auth, query usage, billing, and subscriptions. Use for: login, logout, check usage, view billing summary/breakdown, spending limit, free tier quota, Token Plan status, coding plan status, pay-as-you-go costs, call logs (which requests failed, 4xx/5xx errors, request latency, recent models called, request-id lookup), subscription orders, team seats. Skip for: model browsing, payment methods (use qwencloud-payment)."
---

# QwenCloud Usage

Use this skill for account authentication, usage, billing, call logs, quotas, subscriptions, orders, seats, and the PAYG spending limit. Never fabricate or mock account data.

## Workflow

1. Before running anything, read [references/cli.md](references/cli.md). It is the only source for installation, authentication flow, commands, arguments, pagination, output modes, and exit handling.
2. Read only the result-semantics reference relevant to the request:
   - Usage summary, free tier, PAYG, model breakdown, or request logs: [references/usage.md](references/usage.md)
   - Billing cycles, cost breakdown, or spending limit: [references/billing.md](references/billing.md)
   - Subscription status, orders, or seats: [references/subscriptions.md](references/subscriptions.md)
   - Login, logout, expired credentials, or authentication events: [references/authentication.md](references/authentication.md)
3. For an exact requestId lookup, query directly without adding `--from`, `--to`, or `--period`, or asking the user for a date. Add time filters only when the user explicitly requests them; see [references/cli.md](references/cli.md).
4. Interpret and display the result according to the Output Contract below.

## Output Contract

### Format and Presentation

Follow the output-mode rules in the CLI reference. Present only the human-readable information relevant to the user's question, and place optional analysis after the result.

Do not display Coding Plan in user-facing output. Omit `coding_plan`, `codingPlanStatus`, and their child fields, including "not subscribed" placeholders, even when returned by the CLI.

### Result Rules

- Classify the result as `success`, `partial`, `empty`, `confirmation_required`, or `error`.
- If a top-N limit omits rows, pagination is unfinished, or the CLI reference's integrity checks fail, use `partial`; never imply the data is complete.
- For Free Tier, show every returned model when the user asks for all quotas. If the answer shows only a subset, state the shown/total model count, mark it `partial`, and include a complete command from [references/cli.md](references/cli.md) to view all free-tier quotas. An offer to show more is not a substitute for that command.
- When two supported commands return conflicting values for what appears to be the same record, show the source and scope of each value and mark the result `partial`. Never silently choose one, merge them, or invent a reconciliation.

### Billing and Subscription Data

- Use the currency returned by the operation and never convert it. For the billing-cycle summary only, display its `pretaxAmount`, `tax`, and `aftertaxAmount` values.
- For a cost breakdown, display only the returned `rows[].amount`, the applicable flat-response or slice-level `totalAmount`, and `currency`. For subscription status or orders, display only the amount and currency fields actually present. Never synthesize missing pretax, tax, after-tax, or currency fields.
- In cost breakdowns, omit `groupKey=__tax__` from model rankings and display that row's returned amount separately as tax.
- Calculate available Team seats only from `seatSummary.groups[]`: use `max(seats - assigned, 0)` for each group, then sum them. Do not derive current capacity from the historical seat-instance list.
- Do not recompute returned dates or monetary amounts. Validate the response and apply the presentation rules above.
- When presenting billing or subscription data, always append: “Final amounts are subject to the console billing statement.”

### Product Semantics

- Token Plan quota and balances use decimal Credits. Preserve the full precision returned by JSON for remaining and total Credits, whether encoded as numbers or numeric strings; never round, truncate, or abbreviate them. For example, display `63028.84805624` as `63028.84805624`, not `63,029`. This applies to usage summaries, subscription quota, tiers, seats, and credit packs; there is no per-model Credit breakdown.
- Use exact values from the same quota scope for consumption comparisons. A rounded usage percentage of zero does not prove that no Credits were consumed; do not infer unused status or refund eligibility from rounded displays.
- Coding Plan uses aggregate request counts in `per_5h`, `weekly`, and `monthly` windows; it has no per-model breakdown. If not subscribed, its branch is `{ "subscribed": false }`.
- PAYG exposes total usage only, without an input/output split.
- The per-model breakdown is PAYG-only. A zero or empty breakdown says nothing about Free Tier, Token Plan, or Coding Plan consumption.
- A usage summary has `period: { from, to }` and these product branches:
  - `free_tier[]`: `model_id` plus either `quota: null` or `quota: { remaining, total, unit, used_pct, status, resetDate }`. Only `status=valid` is usable and may contribute to an available-quota total. Treat `expire` as expired and `exhaust` as exhausted even when `remaining` is positive; an unknown status or `quota: null` has unknown availability.
  - `token_plan`: when subscribed, `subscribed`, `planName`, `status`, `totalCredits`, `remainingCredits`, `usedPct`, and `resetDate`; otherwise at least `{ "subscribed": false }`. `resetDate` is the next credit refresh date when present; if absent, report the refresh date as unknown. When subscription auto-renewal is off, note that the plan expires after this date.
  - `coding_plan`: when subscribed, `subscribed`, `plan`, and `windows` containing `per_5h`, `weekly`, and `monthly`, each with `remaining`, `total`, and `used_pct`; otherwise `{ "subscribed": false }`.
  - `pay_as_you_go`: `models[]` entries with `model_id`, `usage`, `cost`, and `currency`, plus `total: { cost, currency }`.

© QwenCloud · Apache-2.0 · QwenCloud/qwencloud-ai

Similar skills

All skills for Agent setup: memory, context, orchestration →

From the same set9

qwencloud-vision A skill for analyzing images and videos with Claude VL and QVQ models: OCR, chart and table reading, multi-image comparison, visual reasoning. CX Codex +1 1,800installs qwencloud-video-generation A skill that generates and edits video with Wan models: text-to-video, image animation, frame transitions. Useful when the agent has to hand back a finished clip. CX Codex +1 1,600installs qwencloud-ops-auth Configures authentication for QwenCloud APIs, checks credentials, and resolves 401 errors. Useful when setting up DASHSCOPE_API_KEY or fixing access issues. CX Codex +1 1,600installs qwencloud-model-selector This skill recommends which Qwen model and parameters fit your task. Handy when you are unsure what is cheaper or what your key can actually reach. CX Codex +1 1,600installs qwencloud-text A skill that routes text generation, chat, code writing and function calling to QwenCloud models. Useful when you want Qwen or third-party models inside your agent. CX Codex +1 1,600installs qwencloud-update-check Checks for updates to the qwencloud-ai skill pack against GitHub releases. It warns you when a newer version is ready to install. CX Codex +1 1,600installs qwencloud-image-generation A skill that generates and edits images with Wan and Claude Image models from inside your agent: text-to-image, style transfer, reference photos. CX Codex +1 1,600installs qwencloud-audio-tts Turns text into speech with Qwen TTS models: voiceovers, narration, audio output for apps. Runs in Claude, Codex, Cursor, Copilot, Gemini CLI and OpenCode. CX Codex +1 1,600installs

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor