← 返回 Skills 市场
ML Pipeline
作者
Dan Repaci
· GitHub ↗
· v0.1.0
635
总下载
0
收藏
4
当前安装
1
版本数
在 OpenClaw 中安装
/install ml-pipeline
功能描述
Complete machine learning pipeline for trading: feature engineering, AutoML, deep learning, and financial RL. Use for automated parameter sweeps, feature cre...
安全使用建议
This package appears coherent with its stated purpose and the included scripts operate on local files (validation, analysis, copying, reporting). Before installing or running: (1) review and run the scripts in a controlled environment (not against sensitive system directories), (2) if you plan to connect to external feature stores or AutoML services, supply credentials only via secure mechanisms and be aware the skill does not declare or manage them, (3) note the deployment script will copy files into whatever target path you provide (use a sandbox or container if unsure). If you need higher assurance, ask the author for provenance or run the code in an isolated VM/container.
功能分析
Type: OpenClaw Skill
Name: ml-pipeline
Version: 0.1.0
The skill is classified as suspicious due to the broad `allowed-tools: Bash` permission declared in `SKILL.md`, which grants the AI agent the ability to execute arbitrary shell commands, posing a significant Remote Code Execution (RCE) vulnerability if the agent is compromised or given malicious instructions. Additionally, the `scripts/pipeline_deployment.py` script performs file system operations (copying and deleting files/directories) based on user-provided paths, which, while necessary for its stated purpose, could be misused for unauthorized file manipulation. There is no clear evidence of intentional malicious behavior within the provided code or documentation, but these capabilities represent significant security risks.
能力评估
Purpose & Capability
The name/description (ML pipeline for trading) aligns with the included scripts and SKILL.md: data validation, feature engineering, AutoML orchestration, evaluation, feature-store integration (described), and deployment. The required resources (none declared) are reasonable for an instruction-only skill containing template scripts.
Instruction Scope
SKILL.md gives high-level, ML-specific guidance (leakage checks, CV, AutoML steps) and asks the agent to gather dataset and pipeline parameters. The bundled scripts operate on local files and directories (validate, analyze, copy, report). There are no instructions to read unrelated system files, harvest environment secrets, or exfiltrate data to external endpoints.
Install Mechanism
No install spec is present (instruction-only skill) and all code is bundled. This is the lower-risk model: nothing is downloaded or exec-installed during install.
Credentials
The skill declares no required env vars or credentials, which matches the bundled scripts (they do local file I/O only). SKILL.md mentions integrating with feature stores (Feast, Tecton) and AutoML libraries — those integrations would typically require credentials or external dependencies, but the skill does not request them. This is not necessarily malicious but means the agent will expect the user to provide any required service credentials or local configs at runtime.
Persistence & Privilege
always is false and the skill does not request persistent system-wide privileges. The deployment script writes to a target directory chosen at runtime and creates a .deployment.json and deployment_report.json in that target — expected behavior for a deployment utility. The skill does not modify other skills or global agent configuration.
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 ml-pipeline - 安装完成后，直接呼叫该 Skill 的名称或使用
/ml-pipeline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
v0.1.0
Major update: ml-pipeline skill is now a unified end-to-end ML pipeline for quantitative trading, consolidating eight previous skills.
- Combines feature engineering, AutoML, deep learning, and financial RL into one authoritative toolkit.
- Provides rigorous anti-leakage validation practices specific to time-series and financial data.
- Includes step-by-step guidance on feature creation, selection, transformation, and evaluation.
- Features templates for pipeline automation including AutoML configuration, cross-validation, and deployment artifacts.
- Covers essential ML concepts (bias-variance, cross-validation) with emphasis on feature engineering decisions.
元数据
常见问题
ML Pipeline 是什么？
Complete machine learning pipeline for trading: feature engineering, AutoML, deep learning, and financial RL. Use for automated parameter sweeps, feature cre... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 635 次。
如何安装 ML Pipeline？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install ml-pipeline」即可一键安装，无需额外配置。
ML Pipeline 是免费的吗？
是的，ML Pipeline 完全免费（开源免费），可自由下载、安装和使用。
ML Pipeline 支持哪些平台？
ML Pipeline 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 ML Pipeline？
由 Dan Repaci（@ahuserious）开发并维护，当前版本 v0.1.0。
推荐 Skills