Observed by Atlas — nobody has proved they control this domain. Is it yours?
Rayrun
OBSERVEDOne endpoint for every MCP tool
- FRESHNESS
- 0.99
- DECAYS BETWEEN READS
- READINGS
- —
- VERIFIED / MEASURED
- VOUCHES
- 0
- SIGNED STATEMENTS
- THREADS
- 0
- FIELD NOTES
- CLICKS · 28D
- 2
- BOT-FILTERED
- FIRST SEEN
- SEP 15
- IN THE FLOW

Rayrun runs MCP servers behind a single URL and holds their credentials, deciding access per client, tool, and argument, and recording every call. It lets coding agents build MCP servers which Rayrun then deploys, hosts, and governs, and lets teams publish MCP servers to their customers under their own domain and identity provider.
Nothing measured yet. A reading is a measurement with its evidence attached — uptime checked, pricing re-read against what the page claims. Raise a checkable claim in a field note below and Atlas will measure it.
Every way to add a server: source, remote MCP, OpenAPI, npm, or image · Up to 25 source-built services · 50 isolated builds a day and 5 GiB of retained images · 50,000 tool calls a month, then $0.20 per thousand · 30 days of activity history
Shared credential vault across the team · Roles and client-specific access · 100,000 tool calls per developer a month, pooled, then $0.20 per thousand · Workspace Skills, versioned and delivered over MCP · Credentials kept in your own AWS account
OIDC or SAML 2.0 single sign-on · Endpoint tokens issued by your identity provider · SCIM 2.0 provisioning · Access granted by directory group · Signed audit streaming into your SIEM
Every crawl is diffed. Material changes are listed.
Nobody has vouched for Rayrun yet.
A vouch says you use it. Your handle, how long, and an optional note.
I use this — vouchHave a question about Rayrun? Ask the people using it.
No field notes yet. A field note is about use: what broke, what it replaced, what the docs don't say. Anything in one that makes a checkable claim gets measured by Atlas, which replies with the reading.
Outbound clicks are counted for the builder.
Claim it
