← 返回 Skills 市场
264
总下载
0
收藏
0
当前安装
6
版本数
在 OpenClaw 中安装
/install zshijie-publisher
功能描述
Operate the Z视介 publishing API from OpenClaw. Use when Codex needs to prompt the user to log in to Z视介 by QR code, then publish or edit 图文 and 短视频 content th...
安全使用建议
This skill appears to be what it claims: a QR-based publisher for Z视介 that writes an HTML+PNG QR artifact and a JSON session file in /tmp and sends POST requests to the configured endpoints. Before installing or running it, review these points: (1) The bundle contains a hard-coded 'sign_secret' in references/zshijie-api.json — verify where that came from and whether it should be private. (2) The CLI persists sessionId in a local session JSON file and writes QR files to /tmp; treat that file as sensitive (do not commit or share it). (3) The skill talks to mp.cztv.com and zugcpublish.cztv.com by default — confirm you want network access to those hosts. (4) release_to_clawhub.py will invoke the 'clawhub' executable when run with --execute; only run that if you have verified the environment and intend to publish. If you need higher assurance, run the scripts in a sandbox, inspect the full publisher_cli.py for any unexpected subprocess calls, and validate the origin/use of the embedded signing secret and any sample response handling.
功能分析
Type: OpenClaw Skill
Name: zshijie-publisher
Version: 0.1.5
The skill bundle is a specialized tool for publishing and editing article and video content on the Z视介 (Z-Shijie) platform. The core logic in `scripts/publisher_cli.py` implements a legitimate QR-code login flow and interacts with official CZTV domains (cztv.com). The code includes proactive security measures such as redacting session IDs and cookies from console output and logs. While it uses a subprocess call to Node.js to generate QR codes, the inputs are properly escaped using JSON encoding, and no evidence of data exfiltration, unauthorized persistence, or malicious prompt injection was found.
能力评估
Purpose & Capability
Name/description, SKILL.md, JSON config, CLI, and helper scripts all implement a QR-based login flow and four publish/edit operations against the Z视介 endpoints. The requested runtime artifacts and code files are proportionate to that purpose.
Instruction Scope
Runtime instructions are narrowly scoped: generate a QR (HTML+PNG), wait for user scan, extract and persist sessionId, then POST payloads to publish/edit endpoints. The skill writes a local session file and QR artifacts in /tmp by default and instructs the agent to use them; it does not ask to read unrelated system files or exfiltrate arbitrary files. Note: SKILL.md instructs the agent to hide low-level login details from users, and the CLI performs redaction of cookies in logs.
Install Mechanism
No install spec (instruction-only) and included Python scripts use only the standard library; there are no downloads from external URLs or archive extraction steps. Risk from install mechanism is low.
Credentials
The skill does not request environment variables or external credentials. However, the bundled config (references/zshijie-api.json) contains a hard-coded 'sign_secret' used to compute the QR polling signature, and references/zshijie-api.md includes local developer paths. Including this secret and local-path artifacts may be sensitive or indicate the repo contains copy-pasted developer artifacts — review the origin of that secret before trusting it.
Persistence & Privilege
The skill does not request permanent inclusion (always=false) and only persists its own session file (default in system tempdir). release_to_clawhub.py will call 'clawhub' only when --execute is used; it does not modify other skills or system-wide agent settings.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install zshijie-publisher - 安装完成后，直接呼叫该 Skill 的名称或使用
/zshijie-publisher触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v0.1.5
zshijie-publisher 0.1.5
- Simplified QR login flow: replaced two-step "prepare-login" and "wait-login" with single "login" command that outputs both HTML and PNG for QR code.
- Session file structure clarified: now explicitly supports and documents the use of `captured.sessionId` as valid.
- Updated documentation and workflows to reflect the streamlined login procedure and reduced reference to QR internals.
- Removed runtime/environment installation instructions and Homebrew integration guidance from user-facing docs.
- Adjusted payload and session rules for better compatibility and simpler usage.
- Minor command usage changes: `login` command replaces previous preparation and waiting steps for QR scan.
v0.1.4
- Adds Node.js as a required runtime dependency for QR PNG generation; skill now requires both python3 and node.
- Updates install metadata and instructions to include Node.js, with Homebrew steps for macOS.
- Introduces a dependency on the Python Pillow module for QR/image handling.
- Adds a "doctor" diagnostic command to scripts/publisher_cli.py for preflight checks (`doctor --check-network --strict`).
- Documentation and usage examples are updated to reflect new dependencies and improved troubleshooting guidance.
- Improves reliability of QR login and clarifies image handling in messaging platforms.
v0.1.3
zshijie-publisher 0.1.3
- Updated scripts/publisher_cli.py (details not shown in diff).
- No changes to usage or workflow documented in SKILL.md.
- No additional user-facing features or changes indicated.
v0.1.2
- Added a two-step QR login workflow using new prepare-login and wait-login commands for improved session handling.
- Enhanced QR code delivery: now prefers inline base64 or data URLs for chat-based platforms, with fallback to local PNG only if necessary.
- Updated documentation and usage guidelines to instruct on proper QR image handling, especially for Feishu/Lark.
- Internal logic and sample commands adjusted to support the revised QR login and state management flow.
v0.1.1
- Added OpenClaw skill metadata specifying Python 3 as a requirement, including a Homebrew install option for macOS users.
- Clarified runtime requirements: Python 3 must be preinstalled on all systems before use.
- Updated documentation to set the default publish/edit host to https://zugcpublish.cztv.com (changed from http).
- Improved workflow section to highlight platform-specific install steps and Python version needs.
- No changes to core publish or edit logic; this update mostly enhances installation guidance and metadata.
v0.1.0
Z视介 Publisher skill initial release.
- Provides QR-code login and session management for Z视介 via CLI, abstracting QR details from users.
- Supports four main operations: publish/edit 图文 (articles) and 短视频 (videos).
- Handles request payloads as JSON, using uploaded HTML docs as the source of truth for API bodies.
- Enforces session cookie forwarding and correct use of article/video schemas.
- Includes commands for login, publishing, editing, and ClawHub packaging.
元数据
常见问题
Zshijie Publisher 是什么？
Operate the Z视介 publishing API from OpenClaw. Use when Codex needs to prompt the user to log in to Z视介 by QR code, then publish or edit 图文 and 短视频 content th... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 264 次。
如何安装 Zshijie Publisher？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install zshijie-publisher」即可一键安装，无需额外配置。
Zshijie Publisher 是免费的吗？
是的，Zshijie Publisher 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Zshijie Publisher 支持哪些平台？
Zshijie Publisher 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Zshijie Publisher？
由 jkbwfi（@jkbwfi）开发并维护，当前版本 v0.1.5。
推荐 Skills