Agent repo

Decomposes the research target into sub-questions and search queries.

Discovers candidate sources across web search connectors.

Runs web searches for the planned queries.

Summarizes recent news and momentum signals for the target.

Identifies competitors and positions them relative to the target.

Scores and ranks gathered sources by trustworthiness and relevance.

Extracts companies, people, products, and technologies as entities.

Reconciles all agent outputs into an executive summary and findings.

Writes the human-readable Markdown research report.

Fetches and extracts content from candidate pages.

Commits the research output to the connected GitHub repository.

Produces an overall score for a repo/target.

Fetches and extracts the full text of the top discovered pages.

Searches SEC EDGAR filings for the target (public companies).

Directional financial read (revenue/margin trends). Not filed data.

Rough valuation framing (comparables, drivers). Not investment advice.

Writes an M&A/investment memo in Markdown.

Searches GitHub for the target's repositories and activity.

Summarizes a repository's README and positioning.

Infers hiring momentum and functional focus from signals.

Assesses market size, growth, and structure.

Builds entity/relationship triples for the knowledge graph.

Finds market reports and landscape sources.

Surfaces key risks with severity and rationale.

Flags conflicting claims across sources.

Validates that claims are backed by cited sources.

Reviews repository license and usage implications.

Discovers and filters to authoritative legal/regulatory sources.

Finds content gaps vs. competitors for SEO.

Clusters keywords into topic groups.

Crawls a domain sitemap to enumerate indexable pages.

Gathers search-engine result pages for target keywords.