Hex Research
Track hypotheses, research runs, source quality, and audits in a graph your agent can update while you validate what to build.
Overview
io.github.levnikolaevich/hex-research-mcp is an Idea-phase MCP server that maintains a research graph of hypotheses, goals, runs, source quality, audits, and generated maps for agents.
What is this MCP server?
- Research graph linking hypotheses, goals, runs, and audit trails
- Source quality signals so agents do not treat every link as equal evidence
- Generated maps summarizing what was learned across runs
- stdio npm @levnikolaevich/hex-research-mcp v0.2.1
- Designed to pair with Hex Graph and Hex Line in agent-heavy workflows
- MCP server version 0.2.1
- npm package @levnikolaevich/hex-research-mcp
- stdio MCP transport
Community signal: 488 GitHub stars.
What problem does it solve?
Founders lose threads when research lives in ephemeral chats, so hypotheses, sources, and audit history never survive into validate-or-build decisions.
Who is it for?
Indie builders running repeated competitive, technical, or audience research who want agent-assisted notes with explicit quality and audit metadata.
Skip if: One-off trivia lookups or pure code editing sessions where Hex Line or Hex Graph is the right tool.
What do I get? / Deliverables
After install, your agent works against a durable research graph so each run, source rating, and audit compounds into maps you can reuse for scoping and docs.
- Persistent research graph with goals, runs, and hypotheses
- Source quality and audit metadata agents can cite
- Generated maps summarizing research progress
Recommended MCP Servers
Journey fit
Idea-phase research is the first shelf for structured inquiry—Hex Research formalizes what solo builders investigate before committing scope. Research subphase fits graph-backed hypotheses, goals, and source scoring rather than one-shot chat answers.
How it compares
Research-graph MCP for hypotheses and sources, not a code editor or self-awareness coaching server.
Common Questions / FAQ
Who is hex-research-mcp for?
It is for solo builders and small teams using AI agents who need structured research—hypotheses, runs, source quality, and audits—before and during product decisions.
When should I use hex-research-mcp?
Use it during Idea research and Validate scoping when you want a graph-backed record of what you tested, which sources mattered, and what maps you generated—not a disposable chat log.
How do I add hex-research-mcp to my agent?
Add @levnikolaevich/hex-research-mcp via npm as a stdio MCP server in your client configuration following the registry entry and repository docs.