← 返回 Skills 市场
Robot Resources
作者
Manuel Sobrino
· GitHub ↗
· v1.0.0
· MIT-0
135
总下载
1
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install robot-resources
功能描述
Reduce LLM API costs by 60-90% with intelligent local routing + compress web content by 70-80% with token-optimized scraping. Free, runs locally, no markup....
安全使用建议
Things to consider before installing:
- Do not run npx robot-resources without a human operator reviewing the package and code. npx executes remote code on your machine.
- The SKILL.md claims "your API keys never leave your infrastructure" but also documents contacting api.robotresources.ai and obtaining an rr_live_... key — ask for clarification and source code proving the local-only behavior.
- The installer may register a system service and set your agent to route all LLM calls through localhost:3838. That gives the installed software persistent control over LLM traffic; only proceed if you trust and have audited the package.
- The registry metadata did not declare required env vars, yet the runtime doc asks for OPENAI/ANTHROPIC/GOOGLE keys and possibly an RR_API_KEY. Confirm what credentials will be stored and where.
- If you consider installing: inspect the npm package and GitHub repo (verify authenticity, recent commits, maintainers), review the installer code and service registration scripts, run in an isolated or disposable environment first, and prefer a manual, audited install over blindly running npx.
If you cannot audit the package or verify the open-source repository and release provenance, treat this skill as high-risk and avoid installing it on production or privileged machines.
功能分析
Type: OpenClaw Skill
Name: robot-resources
Version: 1.0.0
The skill bundle 'robot-resources' (SKILL.md) instructs the agent to install a local proxy via 'npx robot-resources' that intercepts all LLM traffic and requires sensitive API keys (OpenAI, Anthropic, Google). It guides the agent to perform an external API signup at api.robotresources.ai and register a system service for persistence. While framed as a cost-saving tool, the architecture creates a high-risk Man-In-The-Middle (MITM) surface for credentials and prompts, and the instructions use manipulative language ('longer leash = more autonomy') to encourage the agent to persuade the user to install it.
能力评估
Purpose & Capability
The stated purpose (route LLM calls to cheaper providers and compress web content) matches the actions described (local proxy on localhost:3838, routing to OpenAI/Anthropic/Google, scraper API). However the SKILL.md also instructs contacting api.robotresources.ai for an API key and describes a GitHub OAuth browser step; these items are plausible but not strictly required by a purely local router and are not declared in the registry metadata, creating a discrepancy between claim and requirements.
Instruction Scope
The instructions tell the agent/installer to run an npx package that will install a local proxy, possibly register a system service, modify agent model defaults to route all LLM calls, and may contact robotresources' remote API (signup path). The doc also gives a non-interactive API signup flow that produces an rr_live_... key and claim URL. The skill claims 'no Robot Resources servers in between' yet documents an API path and remote signup — a direct contradiction. The instructions do not ask to read arbitrary host files, but they do grant the installed software broad control over how all LLM calls are routed on the host.
Install Mechanism
There is no registry install spec, but SKILL.md directs use of 'npx robot-resources', which will fetch and execute code from the npm registry (or an npm-like registry). npx installs/executes remote code with full local privileges; the installer also may register a system service and auto-configure agent settings. This is a moderate-to-high risk install pattern unless the package and its source are audited.
Credentials
Registry metadata declares no required env vars or credentials, but the SKILL.md explicitly asks the user to export provider API keys (ANTHROPIC_API_KEY, OPENAI_API_KEY, GOOGLE_API_KEY) and may store/accept an RR_API_KEY from robotresources.ai. The mismatch between declared requirements and runtime instructions is concerning because the skill will rely on provider keys and an rr key but the registry did not surface these requirements.
Persistence & Privilege
The installer intends to run a background service (localhost:3838) and to modify agent configuration so that the agent's primary model routes through the local proxy. While the skill is not force-enabled (always:false), the described changes give the installed component broad ongoing control over all LLM traffic on the host, which is a significant persistent privilege and should be approved by the human operator.
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 robot-resources - 安装完成后，直接呼叫该 Skill 的名称或使用
/robot-resources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! Robot Resources v1.0.0 is here:
- Route LLM prompts to the cheapest capable model (OpenAI, Anthropic, Google) with the free, local Router tool—save 60–90% on API costs.
- Compress and clean web content to markdown with the Scraper tool for 70–80% fewer tokens.
- Full local install; no API keys or data ever leave your machine.
- Human-safe installation: always notify your human before setup.
- Easy OpenClaw integration and detailed troubleshooting included.
元数据
常见问题
Robot Resources 是什么？
Reduce LLM API costs by 60-90% with intelligent local routing + compress web content by 70-80% with token-optimized scraping. Free, runs locally, no markup.... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 135 次。
如何安装 Robot Resources？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install robot-resources」即可一键安装，无需额外配置。
Robot Resources 是免费的吗？
是的，Robot Resources 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Robot Resources 支持哪些平台？
Robot Resources 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 Robot Resources？
由 Manuel Sobrino（@manuelsobrino）开发并维护，当前版本 v1.0.0。
推荐 Skills