web-research

Search the web via DuckDuckGo and fetch public URLs with security controls.

3|1|Updated Feb 14, 2026
One-click install
npx skills add https://github.com/LiboMa/agenticops-chat --skill web-research-liboma
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: web-research
Source: https://github.com/LiboMa/agenticops-chat/tree/main/skills/web-research
Command: npx skills add https://github.com/LiboMa/agenticops-chat --skill web-research-liboma

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? During incident investigation, agents often need external information — cloud provider status pages, CVE details, documentation, changelogs — but lack a safe, controlled way to reach the open internet. This Skill provides web search and URL fetching with built-in security guardrails so investigations can incorporate external evidence without exposing internal networks. ## Core Features & Use Cases - Web Search: Query DuckDuckGo and receive titles, URLs, and snippets, then chain promising results into fetches for deeper reading. - Secure URL Fetching: Retrieve public HTTP/HTTPS pages with private-IP blocking, GET/HEAD-only methods, 30s timeouts, and 1MB response limits. - Use Case: When multiple EKS nodes show the same symptoms, search for the error message, fetch the AWS health status page to rule out a provider outage, and pull CVE details from the NVD API to cross-reference security findings. ## Quick Start Activate the web-research skill and search for the error message you are investigating, then fetch the most relevant result URL for details.

Frequently Asked Questions about web-research

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I search the web from an AI agent during incident investigation?▼

Use the web_search tool with a query string and optional max_results (1-20, default 8). It queries DuckDuckGo and returns numbered titles, URLs, and snippets that you can pass to web_fetch for full page content.

How do I check AWS or Azure status pages programmatically?▼

Fetch the provider status URL with web_fetch, such as https://health.aws.amazon.com/health/status for AWS or https://status.azure.com/en-us/status for Azure. JSON endpoints like the AWS status.json are easier to parse than HTML pages.

Can web_fetch access internal or private network URLs?▼

No, web_fetch blocks private IP ranges (10.x, 172.16.x, 192.168.x, 127.x), cloud metadata endpoints (169.254.169.254), and IPv6 loopback. Only public HTTP/HTTPS URLs are allowed; use internal tools like run_on_host or run_kubectl for private data.

How do I look up CVE details from the NVD API?▼

Fetch https://services.nvd.nist.gov/rest/json/cves/2.0?cveId=CVE-YYYY-NNNNN with web_fetch to get JSON containing severity, description, and affected versions. NVD rate-limits to roughly 5 requests per 30 seconds without an API key.

What are the limitations of web_fetch for large pages?▼

Responses are capped at 1MB with a 30-second timeout, and output is truncated to 4000 characters. Redirects are not auto-followed, so you must call web_fetch again with the returned target URL. Only GET and HEAD methods are supported.