← 返回 Skills 市场
530
总下载
0
收藏
2
当前安装
5
版本数
在 OpenClaw 中安装
/install uxc
功能描述
Discover and call remote schema-exposed interfaces with UXC. Use when an agent or skill needs to list operations, inspect operation schemas, and execute Open...
安全使用建议
This skill appears to be what it says: a wrapper for invoking schema-exposed remote APIs via the uxc CLI. Before installing or running anything: (1) verify and review the upstream install script if you intend to use it (don't run unreviewed scripts), (2) prefer Homebrew or cargo if you trust those package sources, (3) be deliberate when configuring credentials—use env variables or a secrets manager (1Password) rather than pasting literals, and (4) remember the skill expects a local uxc daemon/links to be configured so network calls executed by uxc will reach whatever endpoints you bind; only bind credentials to hosts you trust.
功能分析
Type: OpenClaw Skill
Name: uxc
Version: 1.0.4
The uxc skill bundle provides an interface for a 'Universal eXecution Client' CLI tool. It is classified as suspicious due to several high-risk capabilities documented for the AI agent, including instructions for a curl|bash installation method in SKILL.md, extensive credential management involving environment variables and 1Password integration in references/auth-configuration.md, and the ability to create new system commands via 'uxc link' in references/usage-patterns. While these features are aligned with the tool's stated purpose of interacting with various API protocols (OpenAPI, GraphQL, gRPC, MCP), they provide a broad attack surface and potential for environment modification. No clear evidence of malicious intent or data exfiltration was found. Primary external references include github.com/holon-run/uxc.
能力评估
Purpose & Capability
The SKILL.md describes a CLI-based adapter that expects a locally installed 'uxc' (and optionally 'grpcurl') which matches the stated purpose of discovering and calling schema-exposed interfaces. The repository includes a validation script that requires 'rg' (ripgrep) but this is a development/validation helper and not needed for normal runtime use.
Instruction Scope
Runtime guidance confines actions to discovery and invocation of remote schema endpoints via the uxc CLI and to setting up link commands and credentials. The docs reference using environment-backed secrets and 1Password templates for auth, but the skill itself does not demand or attempt to read local secrets; those are configuration patterns for the user. The SKILL.md does recommend running an install script but explicitly advises reviewing it before running.
Install Mechanism
There is no platform install spec (instruction-only skill), so nothing is written to disk by the skill. The docs instruct users to install uxc via Homebrew, cargo, or an install script fetched from raw.githubusercontent.com/holon-run/uxc — pointing to a GitHub raw URL is common, but users should review any fetched script before executing it.
Credentials
The skill declares no required environment variables or credentials. The reference docs describe patterns for using env-backed secrets, 1Password, and OAuth only as configuration options for the uxc runtime; those are appropriate and proportional to a CLI that must authenticate to remote services.
Persistence & Privilege
The skill does not request always: true and does not modify other skills or global agent settings. It encourages creating persistent local link commands via 'uxc link' which is expected behaviour for a wrapper CLI and is limited to the user's environment.
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 uxc - 安装完成后，直接呼叫该 Skill 的名称或使用
/uxc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
v1.0.4
- Clarified and expanded authentication guidance in SKILL.md, directing users to specific reference tracks for API keys, multi-field auth, and OAuth.
- Updated references/auth-configuration.md and references/usage-patterns.md to reflect new authentication approaches.
- Improved reference documentation to clarify handling of request signers and multi-field secrets.
v1.0.3
- Updated authentication documentation to clarify API key configuration (headers/query params, secret sources).
- Improved references in SKILL.md for authentication configuration details.
- Minor wording and clarity improvements in reference sections.
v1.0.2
- Added guidance for creating links to stdio hosts with credential-driven child environment authentication in the link-first wrapper skill workflow.
- Updated equivalence rules to document use of `--credential` and `--inject-env` when invoking links.
- Minor clarifications and expansions to link creation instructions in the SKILL.md documentation.
v1.0.1
- Added reference file for authentication configuration: references/auth-configuration.md
- Updated link workflow to support OpenAPI schema URLs in SKILL.md and usage documentation
- Improved documentation on overriding and persisting schema URLs with link commands
- Enhanced reference file listings to include authentication configuration details
v1.0.0
- Initial release of the uxc skill for discovering and calling remote schema APIs via a single CLI workflow.
- Supports OpenAPI, GraphQL, gRPC Reflection, MCP, and JSON-RPC/OpenRPC interfaces.
- Provides consistent, machine-readable output envelopes for both success and failure cases.
- Documents link-first workflow for wrapper skills and best practices for skill reuse.
- Includes guidance on installation, input modes, authentication, and output parsing.
- Reference files are provided for detailed usage patterns, protocol naming, endpoints, OAuth, and error handling.
元数据
常见问题
Uxc 是什么？
Discover and call remote schema-exposed interfaces with UXC. Use when an agent or skill needs to list operations, inspect operation schemas, and execute Open... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 530 次。
如何安装 Uxc？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install uxc」即可一键安装，无需额外配置。
Uxc 是免费的吗？
是的，Uxc 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Uxc 支持哪些平台？
Uxc 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 Uxc？
由 jolestar（@jolestar）开发并维护，当前版本 v1.0.4。
推荐 Skills