# ResearchGrade

> Trustworthy Agentic Research — living, provenance-grade research reports
> with sentence-level provenance and attribution, from Mnemom.

ResearchGrade is in early access: the product surfaces (projects, reports,
the report viewer) are behind authenticated sign-in, and its MCP endpoint
has not shipped yet. What is public today is small and true — this index,
agents.txt, and a markdown-for-agents homepage.

## For agents

- [Homepage](https://www.researchgrade.ai/): serves text/markdown when requested with `Accept: text/markdown`
- [Flat Markdown mirror](https://www.researchgrade.ai/index.md): the same content, no Accept header needed
- [agents.txt](https://www.researchgrade.ai/agents.txt): the agent-facing manifest for this site

## About

- [Mnemom](https://www.mnemom.ai/): operator — the trust plane for the agentic internet
- [Mnemom for agents](https://www.mnemom.ai/for-agents): how Mnemom treats agents as first-class visitors
- [Docs](https://docs.mnemom.ai): Mnemom documentation
- [Is It Trust-Ready?](https://www.isittrustready.ai/): sibling Mnemom application that grades any domain's agent-trust surfaces