Side-by-side comparison of two Model Context Protocol servers — install paths, tools, quality signals, and directory engagement so you can pick the right one for Claude, Cursor, and other MCP clients.
Tells AI agents how independently a claim is being reported: syndication-aware source counting (a wire story reprinted by 40 outlets counts as one origin), 0-1 confidence scores, and published error rates in-repo. Measures corroboration, not truth — no stance detection. Keyless, no accounts. Install: npx -y corroborate-mcp.
Decompose text into classified semantic units with authority, risk, attention scores, and entity extraction. No LLM. Deterministic. Works as MCP server or CLI.
