← Back to Skills Marketplace
wsjwoods

Qianfan Usage

by woods · GitHub ↗ · v1.0.2 · MIT-0
cross-platform ✓ Security Clean
417
Downloads
0
Stars
0
Active Installs
3
Versions
Install in OpenClaw
/install qianfan-usage
Description
Query Baidu Qianfan Coding Plan usage and quota with automatic login or open the control console via commands.
Usage Guidance
This skill appears to implement what it claims (query Qianfan usage and auto-login) and uses only the Baidu console API and a cookie file in your home directory. Before installing: (1) Confirm you are comfortable storing Baidu session cookies at ~/.baidu-qianfan-auth.json and that the cookie file has appropriate filesystem permissions. (2) Ensure agent-browser is installed from a trusted source (the scripts call it extensively); the package is listed as a dependency but the registry entry has no install step. (3) Note the minor metadata mismatches: registry "Requirements" shows none while _meta.json and SKILL.md require QIANFAN_PHONE — set the env var if you want non-interactive login. (4) Review the included scripts (they are plain Python) if you need higher assurance, or run them in a restricted environment/VM. If you do not want the skill to store session cookies or to automate a login flow, do not install or run it.
Capability Analysis
Type: OpenClaw Skill Name: qianfan-usage Version: 1.0.2 The skill bundle is designed to query Baidu Qianfan Coding Plan usage. It uses a browser automation tool (agent-browser) to handle authentication and stores session cookies locally in `~/.baidu-qianfan-auth.json`. The scripts (qianfan_usage.py, check_quota_v2.py) communicate exclusively with official Baidu endpoints (console.bce.baidu.com) and do not exhibit signs of data exfiltration, malicious execution, or prompt injection. While it handles sensitive session data, its behavior is transparent and strictly aligned with its stated purpose.
Capability Assessment
Purpose & Capability
The name/description (query Qianfan usage and open console) align with the included Python scripts: they use saved cookies, call the Baidu Qianfan API endpoint, and implement an automated login flow via agent-browser. Required functionality (phone for SMS login, cookie storage) is consistent with the stated purpose.
Instruction Scope
SKILL.md and the scripts instruct the agent to read QIANFAN_PHONE (from env or prompt), open pages with agent-browser, capture snapshots, fill inputs, and save cookie state to ~/.baidu-qianfan-auth.json. Those actions are within scope for an auto-login/usage checker. Note: SKILL.md suggests putting QIANFAN_PHONE in ~/.openclaw/workspace/.env while code reads os.environ; the meta.json also marks QIANFAN_PHONE required — minor mismatch between documentation and code/registry metadata. The scripts only reference the user's home cookie file and the Baidu console API; they do not access other system paths or unrelated credentials.
Install Mechanism
There is no install spec in the registry entry (instruction-only), which is low-risk. However the documentation and _meta.json list agent-browser as a dependency and SKILL.md says 'agent-browser (auto install)'. Because no install mechanism is provided, agent-browser must already exist on the environment or be installed manually; this is an inconsistency in packaging/documentation but not an active risk in itself.
Credentials
The only sensitive input required is a phone number for SMS-based login (QIANFAN_PHONE) and the scripts save/read a cookie file in the user's home. There are no requests for unrelated secrets (AWS keys, tokens, etc.). Storing and using cookies for authenticated requests is expected for this purpose. The registry-level 'Requirements' reported as 'none' (top summary) conflicts with _meta.json and SKILL.md which require QIANFAN_PHONE; this is a documentation/metadata mismatch to be aware of.
Persistence & Privilege
The skill does save login state to ~/.baidu-qianfan-auth.json, which is normal for a login helper. It does not request always:true and does not modify other skills or system-wide agent settings. Autonomous invocation (disable-model-invocation=false) is the platform default; it is not flagged by itself.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install qianfan-usage
  3. After installation, invoke the skill by name or use /qianfan-usage
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.2
- Migrated main functionality from shell scripts to Python modules. - Added new Python files: check_quota_v2.py, qianfan.py, qianfan_usage.py. - Removed deprecated files: qianfan-usage.sh and qianfan_api.py. - README and documentation updated to reflect codebase changes.
v1.0.1
- Added README.md file with usage instructions and details.
v1.0.0
Initial release of qianfan-usage. - Provides usage and quota details for Baidu Qianfan Coding Plan via simple commands. - Supports querying 5-hour, weekly, and monthly usage, with automatic login using agent-browser. - Requires phone number configuration via environment variable. - Includes API endpoint information and expected response format. - Supports direct console web access via command.
Metadata
Slug qianfan-usage
Version 1.0.2
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 3
Frequently Asked Questions

What is Qianfan Usage?

Query Baidu Qianfan Coding Plan usage and quota with automatic login or open the control console via commands. It is an AI Agent Skill for Claude Code / OpenClaw, with 417 downloads so far.

How do I install Qianfan Usage?

Run "/install qianfan-usage" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Qianfan Usage free?

Yes, Qianfan Usage is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Qianfan Usage support?

Qianfan Usage is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Qianfan Usage?

It is built and maintained by woods (@wsjwoods); the current version is v1.0.2.

💬 Comments