AgentSkillsCN

agent-zero-deep-research

借助Agent Zero的自主浏览功能,开展多源深度研究

SKILL.md
--- frontmatter
name: agent-zero-deep-research
description: Multi-source deep research using Agent Zero's autonomous browsing
version: 1.0.0
author: agent-zero-module
api_only: true
priority: 8
triggers:
  - "research"
  - "investigate"
  - "deep dive"
  - "find out"
  - "look up"
  - "what do you know about"
allowed_endpoints:
  - "POST /api/agent-zero/research"
  - "POST /api/agent-zero/task"
  - "GET /api/agent-zero/status"
tags:
  - research
  - agent-zero
  - web-browsing

Agent Zero Deep Research

When to Use

User wants thorough, multi-source research on a topic. This delegates to Agent Zero which can autonomously browse the web, read documents, and synthesize findings.

API Endpoints

Submit Research

POST http://agent-zero-bridge:8101/tools/research

json
{"query": "topic to research", "depth": "thorough"}

Check Status

GET http://agent-zero-bridge:8101/tools/status

Behavior

  1. Acknowledge the request: "I'll research that for you..."
  2. Submit to Agent Zero
  3. Wait for results (can take 30-120 seconds)
  4. Present findings in structured format with sources

Important

  • Always tell user research is in progress
  • Cite sources when available
  • For time-sensitive topics, note the date of information