Internet Research Review

Portable agent skill for public internet research. It helps an AI agent find, compare, rank, and summarize current evidence from public sources.

Use Cases

• Source-backed research reports
• Product, company, or market scans
• Technical evidence gathering
• Community pain point and sentiment review
• Competitor and alternative analysis
• News, changelog, incident, and issue review

What It Covers

The skill guides an agent to search across the source types that matter for the request, including official documentation, changelogs, repositories, issue trackers, forums, tutorials, news, reviews, social discussion, standards, and security advisories.

It favors primary sources, recent evidence, direct links, and compact reporting.

Install

Copy this folder into any agent runtime that supports SKILL.md skills:

skills/internet-research-review/
  SKILL.md
  README.md

Invoke

Use the skill name in your agent prompt:

Use $internet-research-review to review current public evidence about <topic>.

Output Shape

The skill asks the agent to return:

• Executive signal: strongest findings
• Evidence table: source, date, signal, relevance, confidence, link
• Patterns: repeated themes, disagreements, outliers
• Risks: uncertainty, missing sources, bias, stale data
• Next searches: only when more research would materially improve confidence

Notes

• Public sources only.
• No paid APIs required.
• No runtime-specific metadata required.
• No private, paywalled, credentialed, or sensitive content without explicit authorization.