MCP server security: a census of every host in the official MCP Registry
Comments on “MCP server security: a census of every host in the official MCP Registry”
Claim-checked. Scored by all three judges, and its checkable claims were compared with the sources it cites. How far stories are checked
Spellwarden, a security vendor, sent one unauthenticated probe to each of 30,402 hosts in the official MCP Registry, calling no tools. Of 26,548 hosts that answered as MCP servers or asked for sign-in, 62% listed tools anonymously, and 9,691 of those lists include a tool not declared read-only. Only 769 hosts report the 2026-07-28 revision; 6 meet its readiness checks. Instruction-like tool text is rare, about one open list in a hundred, but that rests on a 60-host sample. Figures are Spellwarden's own.
Why it ranked
No. 14 of 143 with a score of 0.73. The panel rated it highest on new information (0.92) and substantiated (0.90), lowest on audience breadth (0.61). Placed No. 14, No. 8 by score alone: the fifth Security story of the day (−0.040). The placement terms move stories; they are not part of the score.
- AI judgeJudge 2claude-haiku-5-5
- Large census of MCP registry hosts with clear methodology and reproducible scanner; strong security relevance for builders, though somewhat vendor-published and narrow in scope.
- AI judgeJudge 3gpt-5.6-luna
- A comprehensive, reproducible census of MCP hosts provides unusually strong new security data and actionable guidance for agent builders, though some conclusions rely on self-reported hints, point-in-time probing, and limited sampled review.
Each judge wrote its line after answering the rubric, with the site and the author still hidden. Every answer behind this is in the table under Score.
Score
0.73total under rubric v2.3
- Significance0.67
- Reader interest0.85
- Audience breadth0.61
- Practical relevance0.84
- Substantiated0.90
- Technical depth0.84
- New information0.92
- Counts against: Promotional0.17
- Counts against: Overclaim0.18
Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).
| Question | Judge 1 | Judge 2 | Judge 3 | Mean | Spread | Weight | Contribution (weight times mean) |
|---|---|---|---|---|---|---|---|
| SignificanceHow much does this change what people in tech know or can do? | 0.68 | 0.55 | 0.78 | 0.67 | 0.23 | +0.24 | +0.161 |
| Reader interestWould a well-informed senior engineer or tech founder want to read this today, and likely pass it on? | 0.82 | 0.80 | 0.93 | 0.85 | 0.13 | +0.26 | +0.221 |
| Audience breadthHow many people who build technology does it concern? | 0.56 | 0.60 | 0.68 | 0.61 | 0.12 | +0.13 | +0.080 |
| Practical relevanceHow useful is it to someone who builds software or tech products? | 0.92 | 0.80 | 0.82 | 0.84 | 0.12 | +0.22 | +0.186 |
| SubstantiatedHow well is it backed? Consider whether it is the primary source, and whether its main claims come with data, code, a demo, a reproducible method or citations. | 1.00 | 0.80 | 0.90 | 0.90 | 0.20 | +0.05 | +0.045 |
| Technical depthHow technically substantive is it? | 0.99 | 0.75 | 0.78 | 0.84 | 0.24 | +0.05 | +0.042 |
| New informationDoes it contain information absent from the stories covered in the last 7 days (list provided)? | 0.87 | 0.90 | 0.98 | 0.92 | 0.11 | +0.05 | +0.046 |
| PromotionalIs it mainly marketing with little substance (a press release, ad or sponsored post with no real information)? A launch post that carries real news is not promotional. | 0.05 | 0.35 | 0.12 | 0.17 | 0.30 | −0.19 | −0.033 |
| OverclaimDo the headline or conclusions claim more than the body supports? | 0.21 | 0.25 | 0.08 | 0.18 | 0.17 | −0.11 | −0.020 |
| Sum of contributions | +0.727 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.727 | ||||||
Spread is the highest minus the lowest judge answer. marks a spread above 0.40: the judges disagree, and the story is marked contested.
The badge for this ranking, for the author's page or anyone's: one line of HTML, and it links back here.
<a href="https://rubric.hrih.si/editions/2026-11-11/14/"><img src="https://rubric.hrih.si/badge/2026-11-11/14.svg" alt="Ranked #14 on Rubric, 11 Nov 2026" width="245" height="28"></a>Claim check
Claim check: 2 matches source · 4 no source given · 2 couldn’t check. Each claim is compared with the source the article itself cites; a flag is published only when both checkers agree.
No source givenNumber
Of 26,548 hosts that answered as MCP servers or asked for a sign-in, 16,470 (62%) listed their tools to a client that had not signed in.
The article gives no link or citation for it.
The article says “16,470 (62%) listed their tools to a client that had not signed in”
The source says No source to compare with.
No source givenNumber
Only 769 hosts speak the 2026-07-28 revision of the protocol, and 6 of them also pass every readiness check for it.
The article gives no link or citation for it.
The article says “Only 769 hosts speak the 2026-07-28 revision of the protocol, and 6 of them also pass every readiness check for it”
The source says No source to compare with.
Couldn’t checkQuote
The MCP Registry's maintainers state that the registry does not verify that a remote URL belongs to the publisher, and since 6 November 2026 several entries may list the same URL.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “The project's maintainers state that the registry does not verify that a remote URL belongs to the publisher, and since 6 November 2026 several entries may list the same URL”
The source says No passage could be compared.
Source cited by the article: github.com/modelcontextprotocol/registry/pull/1806
Couldn’t checkVersion
The 2025-11-25 MCP specification says clients and authorization servers SHOULD support CIMD and MAY support DCR for backwards compatibility, and the 2026-07-28 revision deprecates DCR in favour of CIMD.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “The 2025-11-25 specification says clients and authorization servers SHOULD support CIMD and MAY support DCR for backwards compatibility, and the 2026-07-28 revision deprecates DCR in favour of CIMD”
The source says No passage could be compared.
Source cited by the article: modelcontextprotocol.io/specification/2025-11-25/basic/authorization
No source givenGeneralization
Roughly one open tool list in a hundred steers the model towards its tool, and fewer ask it to keep something from the user.
The article gives no link or citation for it.
The article says “roughly one open list in a hundred steers the model, and fewer ask it to keep something from the user”
The source says No source to compare with.
Matches sourceNumber
Knostic found that all 119 hand-checked servers, from 1,862 internet-exposed servers found through Shodan, listed their tools without authentication.
The cited source says the same thing.
The article says “All 119 checked servers listed their tools without authentication”
The source says “We identified a total of 1,862 MCP servers exposed to the internet. From this set, we manually verified a sample of 119. All 119 servers granted access to internal tool…”
Source cited by the article: knostic.ai/blog/mapping-mcp-servers-study
Matches sourceNumber
Zhou et al. found 40.55% of 7,973 live remote servers expose tools without authentication.
The cited source says the same thing.
The article says “40.55% expose tools without authentication; OAuth 30.45%, static token or key 29.00%”
The source says “3,233 (40.55%) of the validated servers expose tool interfaces with no authentication at all, meaning that any client can invoke tools or trigger API requests without presenting credentials”
Source cited by the article: arxiv.org/abs/2605.22333
No source givenFirst, only or fastest
Spellwarden's report is the first to give figures on Client ID Metadata Document support.
The article gives no link or citation for it.
The article says “the first figures we found on Client ID Metadata Document support”
The source says No source to compare with.
Also covered by
No other coverage was clustered with this story.
Elsewhere
Elsewhere: HN 1 pts, not on front page
Recorded for context only, checked 11 Nov 2026, 05:47 UTC. It played no part in the ranking.