browser-research-agent

An AI research agent that searches, reads, summarizes, and cites web sources for repeatable market and repository intelligence.

This repo is inspired by the continuing popularity of browser-use, Firecrawl, Agent-Reach, and web-data tooling for agents.

Goal

Build an agent that can answer research questions with links, dates, source quality labels, and a short confidence note.

First Workflow

query -> search -> open sources -> extract claims -> compare -> cite -> brief

Features To Build

• Source collector for search results and direct URLs.
• Claim extractor with citation tracking.
• Recency filter for fast-moving topics.
• GitHub repository trend analyzer.
• Markdown report output.

Repository Shape

src/
  search/
  extract/
  rank/
  report/
examples/
  trending-ai-agents.md
docs/
  source-quality.md

Milestone 1

• Done: build a CLI command that accepts a topic and sources.
• Done: load source metadata from Markdown front matter.
• Done: produce a cited Markdown report.
• Done: add tests for citation output.

Run It

python -m browser_research_agent.cli "AI agent trends" examples/source-agent-skills.md examples/source-browser-agents.md
python -m browser_research_agent.cli "AI agent trends" examples/source-agent-skills.md examples/source-browser-agents.md --output examples/trending-ai-agents.md

Test It

PYTHONPATH=src python -m pytest

Publish

This package is ready for GitLab Package Registry and PyPI releases. See RELEASE.md.

Trend Notes

• Agents need fresh web context, especially for trend watching.
• The gap is trust: every claim needs a source, date, and confidence label.
• GitHub, Hacker News, Reddit, docs, and product changelogs should be separate source types.