← 返回 Skills 市场
932
总下载
0
收藏
8
当前安装
1
版本数
在 OpenClaw 中安装
/install programming-assistant
功能描述
Assist with programming tasks using Claude CLI. Use when OpenClaw needs to perform coding tasks, create new programs, modify existing code, or get programmin...
安全使用建议
This skill appears to be a straightforward Claude-CLI-based programming helper, but check a few things before installing or running it:
- Ensure you have (and trust) the Claude CLI: SKILL.md expects the 'claude' command and authentication, but the skill metadata doesn't declare that dependency. Install and authenticate Claude CLI separately and verify how it stores credentials.
- Review scripts/init_project.sh before running: it will create directories/files in whatever target path you pass and will run 'git init' if git is available. Run it in a sandbox or on a safe path first.
- Confirm that you will not inadvertently provide sensitive credentials to the skill or to the Claude CLI session. The skill mentions authentication but does not define an auth mechanism; do not paste secrets into an automated session unless you understand where they go.
- Because the skill can read/write repository files, consider running it with limited filesystem permissions or in a disposable environment if you are unsure.
If you want higher assurance, ask the publisher to update the skill metadata to declare the 'claude' binary as a required dependency and to document exactly how authentication is expected to be provided.
功能分析
Type: OpenClaw Skill
Name: programming-assistant
Version: 1.0.0
The skill is classified as suspicious due to the explicit declaration in `SKILL.md` that the agent has access to 'Bash: For running commands, tests, and system operations'. This capability, while potentially necessary for a programming assistant, introduces a significant risk of Remote Code Execution (RCE) via prompt injection if the AI agent's inputs are not rigorously sanitized. Furthermore, the `scripts/init_project.sh` script takes user-controlled directory paths, which, when combined with Bash access, could allow an attacker to create or modify files in arbitrary locations, leading to file system manipulation or further RCE opportunities. There is no clear evidence of intentional malicious design, but the broad, powerful capabilities present a high vulnerability risk.
能力评估
Purpose & Capability
The SKILL.md explicitly requires Claude CLI (the 'claude' command) and references IDE integration, authentication, and using bash/Edit/Read tools, but the skill metadata lists no required binaries or primary credential. This omission is a coherence gap: a skill that depends on an external CLI should declare that dependency so callers and sandboxing can enforce it. Other than that omission, the provided scripts and docs align with the stated programming-assistant purpose.
Instruction Scope
The instructions direct the agent to read and write files in the target project directory, use the Claude CLI, run shell commands, and review/test generated code. These actions are within the stated purpose (project setup, code generation, editing, review). The SKILL.md does not instruct the agent to access unrelated system areas, exfiltrate data, or contact unknown external endpoints.
Install Mechanism
There is no install spec (instruction-only) and only one small helper script is included. No downloads, package installs, or extraction steps are present. This is the lower-risk model for skills.
Credentials
The skill declares no required environment variables or credentials, which is consistent with the included files. However, the runtime instructions mention that Claude CLI must be 'available and authenticated' — the skill does not declare how authentication should be supplied (env vars, config files, or interactive auth). Confirm how the Claude CLI will be authenticated before use; do not store credentials inside this skill.
Persistence & Privilege
The skill is not marked always:true and does not request system-wide configuration changes. The included script will create project directories and (if git is installed) run 'git init' and create a .gitignore — these are expected behaviors for a project-initialization helper but do perform filesystem changes in the target directory. The skill does not attempt to modify other skills or global 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 programming-assistant - 安装完成后，直接呼叫该 Skill 的名称或使用
/programming-assistant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
v1.0.0
- Initial release of the Programming Assistant skill.
- Provides structured workflows for project setup, code generation, modification, debugging, and code review using Claude CLI.
- Supports seamless integration with Claude CLI tools for coding, file editing, and project management.
- Includes step-by-step usage patterns for common programming tasks.
- Offers guidance on prerequisites, error handling, and best practices for secure, maintainable code.
元数据
常见问题
Programming Assistant 是什么？
Assist with programming tasks using Claude CLI. Use when OpenClaw needs to perform coding tasks, create new programs, modify existing code, or get programmin... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 932 次。
如何安装 Programming Assistant？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install programming-assistant」即可一键安装，无需额外配置。
Programming Assistant 是免费的吗？
是的，Programming Assistant 完全免费（开源免费），可自由下载、安装和使用。
Programming Assistant 支持哪些平台？
Programming Assistant 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 Programming Assistant？
由 Leo Wing（@leowing）开发并维护，当前版本 v1.0.0。
推荐 Skills