← 返回 Skills 市场

Research Assistant

Name: Research Assistant
Author: banxian87

作者 banxian87 · GitHub ↗ · v1.0.0 · MIT-0

cross-platform ⚠ suspicious

110

总下载

当前安装

版本数

在 OpenClaw 中安装

/install ai-research-pro

功能描述

Research assistant using ReAct + Plan-and-Solve for web research, information synthesis, and report generation with citations.

使用说明 (SKILL.md)

Research Assistant

AI-powered research assistant that combines tool use (ReAct) with structured planning (Plan-and-Solve) to conduct thorough research and generate comprehensive reports.

Features

🔍 Web Research

Search Queries: Intelligent search strategy
Source Evaluation: Credibility assessment
Information Extraction: Key facts and data

📊 Analysis & Synthesis

Multi-perspective: Compare different sources
Fact-checking: Verify claims across sources
Trend Analysis: Identify patterns

📝 Report Generation

Structured Reports: Clear organization
Citations: Proper source attribution
Executive Summary: Key findings upfront

Usage

const researcher = new ResearchAssistant();

const report = await researcher.research({
  topic: 'Impact of AI on Software Development Jobs',
  depth: 'comprehensive',  // quick, standard, comprehensive
  sources: 10,  // minimum sources
  includeStats: true
});

console.log(report.summary);
console.log(report.citations);

Architecture

User Request
    ↓
Plan-and-Solve Agent
    ├─ Phase 1: Define research questions
    ├─ Phase 2: Search strategy
    ├─ Phase 3: Source collection
    ├─ Phase 4: Analysis
    └─ Phase 5: Report writing
    ↓
ReAct Agent (for each phase)
    ├─ Search web
    ├─ Fetch pages
    ├─ Extract info
    └─ Verify facts
    ↓
Final Report with Citations

Installation

clawhub install research-assistant

License

MIT

Version

1.0.0

Created

2026-04-02

安全使用建议

This package appears to be a stub: it documents ReAct-based web research but the code returns simulated sources (example.com) and a placeholder LLM response. Before installing or relying on it: 1) confirm whether you need a real web-search/HTML-fetch implementation — this version does not perform network fetching; 2) if you expect real searches, request the author to show where/how search APIs and credentials would be used; 3) do not trust the generated 'citations' from this version — they are synthetic; 4) if a later version adds network calls or credentials, review those changes carefully (endpoints, required env vars, and install steps) and test in a sandbox. If you only need a local prompt-driven report generator (no live web access), this code may be acceptable but limited.

功能分析

Type: OpenClaw Skill Name: ai-research-pro Version: 1.0.0 The skill bundle implements a standard research assistant template using a Plan-and-Solve architecture. The code in index.js consists of boilerplate logic for prompt construction and a simulated search function, with no evidence of data exfiltration, malicious execution, or harmful prompt injection.

能力评估

ℹ Purpose & Capability

Name/description promise web research with tool use (ReAct) and Plan-and-Solve. The supplied code implements planning and prompt-driven synthesis but does not perform real web searches, page fetches, or tool integration — it returns simulated sources (https://example.com/...) and uses a default LLM stub. This mismatch suggests the package is a stub or incomplete rather than actually providing the advertised web-research capability.

ℹ Instruction Scope

SKILL.md instructs the agent to 'Search web', 'Fetch pages', 'Extract info', and run ReAct loops. The runtime instructions in SKILL.md are consistent with a web-research tool, but the code does not implement those actions and instead simulates results. The documentation could lead a user to expect network activity and real citations that this version does not perform.

✓ Install Mechanism

No install spec and no downloads; the skill is instruction-plus-source only. Nothing is written to disk by an installer, which reduces install-time risk.

✓ Credentials

The skill requests no environment variables, credentials, or config paths. That is proportionate to the included implementation (which performs no network I/O or external API calls).

✓ Persistence & Privilege

always is false and there is no mechanism shown to modify agent-wide configs or persist credentials. The skill does not request elevated persistence or privileges.

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 ai-research-pro
安装完成后，直接呼叫该 Skill 的名称或使用 /ai-research-pro 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

AI research assistant with ReAct + Plan-and-Solve

元数据

Slug ai-research-pro

版本 1.0.0

许可证 MIT-0

累计安装 0

当前安装数 0

历史版本数 1

常见问题