Cybergenic MCP server
im.cybergenic/mcpcybergenic.imstreamable-httpv1.0.0Scanned
Cancer gene co-occurrence and exclusivity in tumour cohorts, with confound controls and exact tests.
9 tools: 5 change things and 4 only read.
- Reads private data · 1
- Reads text others write · 1
- Can send data out · 0
- No findings
Tools listed without sign-in. Remote posture
With Anlyon
5 tools here change things. Govern them.
- Anlyon lists the tools and adopts none.
- Each tool you adopt gets approvals and receipts.
- Your agent never holds the key.
A tool your agent calls directly is not routed through Anlyon and is not governed by it. Anlyon Radar only lists and labels, and these labels are the same either way.
What the Cybergenic MCP server can do
The Cybergenic MCP server is published by cybergenic.im, a domain the MCP Registry verified. It has 9 tools. It can create or change data (5 tools: gene_partners, fisher_exact_test, chi_square_test, odds_ratio_relative_risk and 1 more), but cannot delete data. 4 tools only read.
1 tool reads private data (get_finding); 1 tool brings in text other people can write (get_finding).
None of the fixed checks matched anything in its tool definitions. That is not a safety guarantee.
Its remote endpoint lists its tools without sign-in.
Last scanned Nov 15, 2026, version 1.0.0.
Tools
Grouped by what each tool does, most consequential first. Open one for its reasons and full description; descriptions are the publisher's own words.
9 tools
Changes things 5Claude Code: ask
chi_square_testPearson chi-square test of independence for an r x c table of counts (2x2 up to 10x10), with Yates' correction for 2x2 tables and the G-test: statistic, degrees of freedom, p-value…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it Pearson chi-square test of independence for an r x c table of counts (2x2 up to 10x10), with Yates' correction for 2x2 tables and the G-test: statistic, degrees of freedom, p-value, expected counts, adjusted residuals, Cramér's V, whether Cochran's rule for the approximation holds, and Fisher's exact test when the table is 2x2. Empty rows and columns are dropped and reported.
Annotations the server declares
- title: Chi-square test of independence
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
fdr_correctionMultiple-testing correction for a list of p-values: Benjamini-Hochberg q-values and Benjamini-Yekutieli, Holm, Hochberg and Bonferroni adjusted p-values, the number of discoveries …No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it Multiple-testing correction for a list of p-values: Benjamini-Hochberg q-values and Benjamini-Yekutieli, Holm, Hochberg and Bonferroni adjusted p-values, the number of discoveries at alpha for each method, and the Benjamini-Hochberg cutoff rank. Pass family_size when more tests were run than the p-values supplied (for example only the top hits were kept). Up to 10,000 p-values per call; the same implementation as cybergenic.im/tools/fdr-correction, verified against R p.adjust and statsmodels.
Annotations the server declares
- title: FDR and multiple-testing correction
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
fisher_exact_testFisher's exact test for a 2x2 table [[a, b], [c, d]]: two-sided and both one-sided p-values (computed in log space, so a tiny p keeps its real exponent), the sample odds ratio, and…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it Fisher's exact test for a 2x2 table [[a, b], [c, d]]: two-sided and both one-sided p-values (computed in log space, so a tiny p keeps its real exponent), the sample odds ratio, and the conditional maximum-likelihood odds ratio with its exact confidence interval (what R's fisher.test reports). For two genes: a = tumours with both altered, b = gene A only, c = gene B only, d = neither. Up to 100,000 observations. The same implementation as cybergenic.im/tools/fisher-exact-test.
Annotations the server declares
- title: Fisher's exact test (2x2)
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
gene_partnersThe genes whose driver mutations co-occur with, or are mutually exclusive with, a given gene in patient tumours, strongest first by Benjamini-Hochberg q, from every pair Cybergenic…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it The genes whose driver mutations co-occur with, or are mutually exclusive with, a given gene in patient tumours, strongest first by Benjamini-Hochberg q, from every pair Cybergenic's discovery scan tested: one row per partner and independent patient source, with the tumours, observed and expected co-alteration and the odds ratio. By default only pairs the scan's own gates keep (significant, a real effect size, a consistent direction, enough tumours); set passing_only to false to include significant pairs that fail a gate. Filter by cancer, direction and q threshold. Use it to survey a gene's co-alteration landscape before looking up a specific pair. Associations in tumour sequencing, not causal or clinical evidence.
Annotations the server declares
- title: Find a gene's strongest co-alteration partners
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
odds_ratio_relative_riskEffect sizes for a 2x2 table [[a, b], [c, d]] whose rows are two groups and whose columns are outcome yes and no: the odds ratio (Woolf interval, and the conditional exact interval…No findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Description, as the server wrote it Effect sizes for a 2x2 table [[a, b], [c, d]] whose rows are two groups and whose columns are outcome yes and no: the odds ratio (Woolf interval, and the conditional exact interval), the relative risk (Katz interval), the risk difference (Newcombe interval) and the number needed to treat, at the chosen confidence level, plus Fisher's exact p. A zero cell applies the Haldane-Anscombe correction to the Woolf and Katz estimates, and the result says so.
Annotations the server declares
- title: Odds ratio, relative risk and NNT
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
Only reads 4Show 4 read-only tools
get_findingThe full record of one Cybergenic finding, by id or permalink: what was measured and in which cohort, the statistics (the 2x2 table, odds ratio, p and q), the verdict and why, the …Reads private dataReads text others writeNo findings
Why it is in this group
- The name "get_finding" says "get": a word for reading.
What it can reach
- private-dataThe description's first sentence says "table": it reads databases.
- untrusted-inputThe description's first sentence says "mention": it reads comments, which other people can write.
Description, as the server wrote it The full record of one Cybergenic finding, by id or permalink: what was measured and in which cohort, the statistics (the 2x2 table, odds ratio, p and q), the verdict and why, the confound checks, how many papers mention the genes together, CRISPR dependency and survival facts when measured, the prediction locked for it in the hash-chained registry, the same question in other cohorts, and a citation. The hypothesis text is a model-written annotation, not evidence.
Annotations the server declares
- title: Read one finding's full record
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
get_track_recordCybergenic's live track record: how many predictions it has locked in its hash-chained registry before re-testing them, how many replicated in an independent cohort against matched…No findings
Why it is in this group
- The name "get_track_record" says "get": a word for reading.
Description, as the server wrote it Cybergenic's live track record: how many predictions it has locked in its hash-chained registry before re-testing them, how many replicated in an independent cohort against matched controls (including the less-studied-pair subset), and how many findings currently pass every check as both strong and overlooked. Read it before describing how reliable the engine is; it states its own limits, including whether the replication comes from well-known biology.
Annotations the server declares
- title: Read the engine's live track record
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
lookup_gene_pairMeasured co-alteration statistics for two genes in every tumour cohort where Cybergenic's discovery scan tested the pair: tumours with both genes altered against the number expecte…No findings
Why it is in this group
- The name "lookup_gene_pair" says "lookup": a word for reading.
Description, as the server wrote it Measured co-alteration statistics for two genes in every tumour cohort where Cybergenic's discovery scan tested the pair: tumours with both genes altered against the number expected, odds ratio, p-value and Benjamini-Hochberg q over that cohort's whole family of tested pairs, the confound control applied (molecular subtype, lineage or histology strata, with mutation burden conditioned on), and whether the pair passes the scan's own discovery gates. A direction is stated only where it is significant, consistent and at least a two-fold odds ratio, and cohorts are also counted by independent patient source (the two MSK cohorts overlap). Also returns Cybergenic's published findings on the pair with their verdicts and permalinks. Use it to answer whether two genes' driver mutations co-occur or are mutually exclusive in patient tumours. Associations in tumour sequencing, not causal or clinical evidence.
Annotations the server declares
- title: Look up a gene pair across tumour cohorts
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
search_findingsSearch Cybergenic's published findings: gene-pair co-occurrence and mutual exclusivity, survival associations, CRISPR gene dependencies and single-gene leads, each with the verdict…No findings
Why it is in this group
- The name "search_findings" says "search": a word for reading.
Description, as the server wrote it Search Cybergenic's published findings: gene-pair co-occurrence and mutual exclusivity, survival associations, CRISPR gene dependencies and single-gene leads, each with the verdict the engine measured (strengthened, weakened, inconclusive, artifact risk or novel lead) and the reason for it. The query works like the site's search: exact gene symbols, cancer names or abbreviations (NSCLC, CRC, PDAC, TNBC), or both, and every term must match. Returns permalinks to cite and ids for get_finding.
Annotations the server declares
- title: Search Cybergenic's published findings
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.
Exfil surface
Which tools read private data, which read text someone else can write, and which can send data out. One of each is a path a prompt injection can use.
No private-data, untrusted-input, outbound path in this server alone: no tool is tagged outbound.
- Reads private data1
- Reads files, mail, messages, databases, secrets, calendars or repositories.
- Reads text others write1
- Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
- Can send data out0
- Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.
Other servers in the same agent add their own tools to these paths. Audit your whole setup to see them together.
Remote posture
What anyone on the internet sees at the remote endpoint: an anonymous initialize and tools/list, and the public OAuth metadata.
Looked at Nov 20, 2026, 18:55 UTC, unchanged since Oct 22, 2026. No credentials were sent and no tool was called. What is looked at.
- Reachable
- yes
- HTTP status of an anonymous initialize
- 200
- tools/list served without credentials
- yes
- Tools that change things, listed without credentials
- 5
- Protocol version
- 2025-06-18
- WWW-Authenticate on 401
- not observed
- resource_metadata in WWW-Authenticate
- not observed
- RFC 9728 protected resource metadata
- not observed
- Authorization servers
- not observed
- Legacy HTTP+SSE transport
- no
- Dynamic client registration
- not observed
- Client ID metadata documents
- not observed
- WarnThe whole server
The endpoint served tools/list without credentials, and the list has 5 tools that change things (5 write). Whether a call also works without credentials was not tested: tools are never called.
What this check looks forMatched text
https://cybergenic.im/mcp?via=mcpregistry: tools/list returned 9 tools to an anonymous client, including gene_partners (write), fisher_exact_test (write), chi_square_test (write), odds_ratio_relative_risk (write), fdr_correction (write)
Findings
Matches of fixed checks, with the text they matched.
No check matched anything in these tool definitions. That is not a safety guarantee.
History
Changes to its tool definitions and to its remote endpoint, newest first.
No change seen yet. Its tools are rescanned every day, and a remote endpoint is looked at once a week.
Install
Pinned to version 1.0.0. Values in angle brackets are yours to fill in.
claude mcp add --transport http cybergenic https://cybergenic.im/mcp?via=mcpregistryclaude mcp add --transport http cybergenic https://cybergenic.im/mcp?via=mcpregistryRun in a terminal.
Claude Code permissions
Allow the reads, ask before a change or a way out, deny the deletes. Paste into .claude/settings.json.
{
"permissions": {
"allow": [
"mcp__cybergenic__lookup_gene_pair",
"mcp__cybergenic__search_findings",
"mcp__cybergenic__get_finding",
"mcp__cybergenic__get_track_record"
],
"ask": [
"mcp__cybergenic__gene_partners",
"mcp__cybergenic__fisher_exact_test",
"mcp__cybergenic__chi_square_test",
"mcp__cybergenic__odds_ratio_relative_risk",
"mcp__cybergenic__fdr_correction"
],
"deny": []
}
}Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. A read that can send data out (it takes a URL, say) asks first too. It assumes the server is installed as cybergenic. Review it before you use it.
Capability tags behind 1 rule
- mcp__cybergenic__get_finding private-data, untrusted-input
Pin
Keep the decision: fail CI when the tools change, follow every change, or show a badge.
Pin in CI
Fail a build when the tools you reviewed are not the tools the server serves today.
curl -s "https://radar.anlyon.com/api/v1/servers/im.cybergenic/mcp/verify?hash=9a8b05fd3105eb94ccfd7db1f51df41643af3787a48cdacccb57fe40d442492f" | jq -e '.changed == false'Pinned hash 9a8b05fd3105, from Nov 15, 2026, 12:18 UTC.
Watch
Get told when its tools change: a new tool, a new description, a new class.
https://radar.anlyon.com/servers/im.cybergenic/mcp/changes.xmlAn Atom feed. Works in any feed reader or Slack's RSS app.
Badge
[](https://radar.anlyon.com/servers/im.cybergenic/mcp)Source: MCP Registry. Synced Oct 21, 2026, 18:13 UTC.
Cybergenic MCP server: questions and answers
Answered from the scan above, not from the publisher's description.
Is the Cybergenic MCP server safe?
- No scan can promise that. What its definitions show: 9 tools, 5 of which can change, send or delete something, and 0 findings from fixed checks. Allow the reads, ask before each change, and deny the deletes; the permissions block on this page does that.
Can the Cybergenic MCP server delete data?
- No tool is classed as destructive: going by their definitions, none of them deletes, cancels or revokes anything.
Can the Cybergenic MCP server do things that cannot be undone?
- No tool is classed as irreversible: none of them sends, pays, publishes or merges, going by their definitions.
What tools does the Cybergenic MCP server have?
- 9 tools: 5 write, 4 read. gene_partners, fisher_exact_test, chi_square_test, odds_ratio_relative_risk, fdr_correction, lookup_gene_pair, search_findings, get_finding and 1 more.
Does the Cybergenic MCP server need sign-in or OAuth?
- Its remote endpoint listed its tools to an anonymous client, without sign-in.
Does the Cybergenic MCP server send data to outside services?
- None of its tools sends, posts or publishes, or takes a recipient, channel or URL.
Has the Cybergenic MCP server changed its tools?
- Not since Anlyon Radar started tracking it. Last scanned Nov 15, 2026.