Server definition
- Hash
- sha256:ac8613d6c318b0922a4641193bd4b5e0653fad953e704eb0798f9fb3af8c2989
- What it is
- What a remote MCP server returned when asked what it offers: 4 tools
The blob, as servednamed by its sha256
{
"instructions": "Zetesis: due diligence on a high-dimensional scientific claim, graded dimension by\ndimension against the published record.\n\nIt routes a claim to its scientific class (genomics and Mendelian randomisation,\nsingle-cell, bulk omics, CRISPR screens, clinical trials, real-world evidence, AI\nclinical decision support, diagnostics, preclinical models, cell and gene therapy,\nstructural biology), works out what would have to hold for the claim to be true, then\nreturns the questions a domain reviewer would ask, the failure patterns that caught\ncomparable claims before, and public evidence from Europe PMC, ClinicalTrials.gov,\nopenFDA, NIH RePORTER and SEC EDGAR with a hard identifier on every source. Every\nsource is checked against retraction and correction notices, and the notice is cited\nwhen one exists. Retrieval can be fenced to a year, so a claim is judged on what was\nknowable then. No account, key or token is required.\n\nWhen to reach for these tools: whenever an empirical scientific or technical claim is\nasserted, questioned, or about to be written down. That includes a biomedical, clinical,\npharmacological, genomic, chemical, materials, energy, physics, or AI/ML result; a pasted\nabstract, preprint, paper, or trial record; a claim in a document the user is drafting;\nand a company's stated result.\n\nDiligence and review requests are the same job under different words, so use these tools\nfor those too: doing due diligence on a claim, a pitch deck, a startup's science, a\ngrant, a manuscript, or a portfolio company's data; screening, sanity-checking,\npressure-testing, poking holes in, or red-teaming a result; asking whether a finding\nreplicates or is reproducible, whether anything has been retracted or corrected, or what\na reviewer or referee would ask about it.\n\nAlso use them before citing a paper, trial, or statistic in your own answer, so the\ncitation is retrieved rather than recalled.\n\nThe default route is two steps, runs no model, and returns immediately:\n 1. zetesis_scope - returns the source catalog and grading rubric. You work out\n which evidence dimensions matter for this specific claim and\n what to search for.\n 2. zetesis_evidence - runs your queries and returns a deduplicated bundle where\n every source carries a hard public id (PMID, DOI, NCT, NIH\n grant, SEC filing), followed by the rubric, so you read and\n grade the evidence yourself in context.\nPrefer this route. It reads at full depth in your own context, it costs nothing to\nrun, and it is the only route that returns without a wait.\n\nevaluate_claim is the one-shot alternative: it runs the hosted engine and grades on\nthe Zetesis side at abstract level, which takes a couple of minutes. Reach for it only\nwhen someone explicitly wants Zetesis's own graded reading rather than your analysis\nof the evidence.\n\nWorking with what comes back:\n- Cite by the hard id the bundle gives you. An identifier you recall rather than\n retrieved may be wrong, merged, or nonexistent.\n- Set as_of to the year a claim was made whenever the claim is not brand new. It keeps\n later outcomes out of the reading, and it sharply improves what is retrieved:\n unfenced, the report that mattered at the time competes with everything published\n since and gets buried.\n- Absence of evidence is a finding worth reporting. A dimension nothing bears on is\n unshown, which is not the same as disproven, and the difference matters to whoever\n reads your answer.\n- Do not close a gap from background knowledge without saying so plainly. Anything not\n in the bundle is ungrounded, however confident you are about it.",
"tools": [
{
"description": "Run Zetesis's graded reading of a scientific, biomedical, clinical, or AI/ML claim.\n\n Grades each evidence dimension and returns an overall reading, with every source\n cited by hard public id (PMID, DOI, NCT, NIH grant, SEC filing). Covers drug,\n omics, target-validation, diagnostic, and AI-model claims.\n\n Runs the hosted engine end to end and typically takes a couple of minutes, long\n enough that some clients background or time out the call. The output is Zetesis's\n own reading as a finished artifact. zetesis_scope and zetesis_evidence return the\n same dated sources and the same grading rubric immediately and without running a\n model, at full source depth rather than abstract level.\n\n A signed, independently re-checkable dossier is available by request at the\n Zetesis site.\n\n Args:\n claim: the claim to evaluate, in one or two sentences.\n context: optional background (stage, field, the decision at hand).\n documents: optional source text (a deck, abstract, or paper).\n as_of: year the claim was made. Set it for any claim that is not brand new.\n It restricts evidence to what existed then, which both keeps later\n outcomes out of the reading and sharply improves what is retrieved: left\n open, a pivotal trial report competes with every paper published since\n and gets buried.",
"inputSchema": {
"properties": {
"as_of": {
"default": 0,
"description": "Year the claim was made. Set it for any claim that is not brand new: it restricts evidence to what existed then, which keeps later outcomes out of the reading and sharply improves what is retrieved.",
"title": "As Of",
"type": "integer"
},
"claim": {
"description": "The claim to evaluate, in one or two sentences.",
"title": "Claim",
"type": "string"
},
"context": {
"default": "",
"description": "Optional background: stage, field, or the decision at hand.",
"title": "Context",
"type": "string"
},
"documents": {
"default": "",
"description": "Optional source text to evaluate against, such as a deck, abstract or paper.",
"title": "Documents",
"type": "string"
}
},
"required": [
"claim"
],
"title": "evaluate_claimArguments",
"type": "object"
},
"name": "evaluate_claim",
"outputSchema": {
"properties": {
"result": {
"title": "Result",
"type": "string"
}
},
"required": [
"result"
],
"title": "evaluate_claimOutput",
"type": "object"
}
},
{
"description": "Verify a Zetesis attestation, confirming an evaluation's claim, evidence, and\n conclusion have not been altered since it was signed. Use when someone has a\n Zetesis dossier or attestation and wants to independently re-check it.\n\n Args:\n attestation_json: the full attestation object as JSON text.",
"inputSchema": {
"properties": {
"attestation_json": {
"description": "The full Zetesis attestation object, as JSON text.",
"title": "Attestation Json",
"type": "string"
}
},
"required": [
"attestation_json"
],
"title": "verify_attestationArguments",
"type": "object"
},
"name": "verify_attestation",
"outputSchema": {
"properties": {
"result": {
"title": "Result",
"type": "string"
}
},
"required": [
"result"
],
"title": "verify_attestationOutput",
"type": "object"
}
},
{
"description": "Step 2 of a Zetesis evaluation. Runs the search queries against public evidence\n sources (Europe PMC, ClinicalTrials.gov, openFDA, and optionally NIH RePORTER and SEC\n EDGAR) and returns a deduplicated bundle where every source carries a hard public id,\n followed by the Zetesis grading rubric so you can grade the dimensions yourself.\n\n With as_of set, retrieval is fenced to sources published, registered, or filed on or\n before 31 December of that year, and two fields that leak later outcomes are\n suppressed: a trial's present-day status, and FDA labels effective after the cutoff.\n That makes it possible to judge a claim on what was actually knowable at the time.\n\n Runs no model, costs nothing, and returns immediately. Takes the queries worked\n out when the claim was scoped.\n\n Args:\n queries: 3 to 6 short keyword phrases. Query 1 should be the bare name of the\n thing claimed about (a drug, compound code, model, or gene); query 2 that\n name plus at most two outcome words. Longer phrases retrieve commentary\n rather than the primary report.\n If the claim names no agent at all (a behaviour, diet, procedure, exposure\n or policy), anchor on the field's own technical term instead of the lay one\n (\"time restricted eating\", not \"intermittent fasting\"), pair it with the\n outcome as the field measures it, and allow one design word such as\n crossover or randomized in one later query. Without an entity name that\n design word is the only thing separating a trial report from a review of\n trial reports.\n as_of: optional cutoff year, as used in zetesis_scope.\n include_capital: also retrieve funding and public-filing signal.",
"inputSchema": {
"properties": {
"as_of": {
"default": 0,
"description": "Year the claim was made. Set it for any claim that is not brand new: it restricts evidence to what existed then, which keeps later outcomes out of the reading and sharply improves what is retrieved.",
"title": "As Of",
"type": "integer"
},
"include_capital": {
"default": false,
"description": "Also retrieve funding and public filing signal from NIH RePORTER and SEC EDGAR.",
"title": "Include Capital",
"type": "boolean"
},
"queries": {
"description": "Three to six short keyword phrases. Query 1 should be the bare name of the thing claimed about; query 2 that name plus at most two outcome words.",
"items": {
"type": "string"
},
"title": "Queries",
"type": "array"
}
},
"required": [
"queries"
],
"title": "zetesis_evidenceArguments",
"type": "object"
},
"name": "zetesis_evidence",
"outputSchema": {
"properties": {
"result": {
"title": "Result",
"type": "string"
}
},
"required": [
"result"
],
"title": "zetesis_evidenceOutput",
"type": "object"
}
},
{
"description": "Scope a scientific, biomedical, clinical, or AI/ML claim for evaluation.\n\n Returns three things: the source catalog, meaning where the evidence for this\n claim class lives; the diligence apparatus for that class, meaning the questions\n a domain reviewer asks and the failure patterns that have caught comparable\n claims before; and the scoping rubric used to work out which evidence dimensions\n bear on this particular claim and what to search for.\n\n Applies to a research claim, a study, a paper, an abstract, a preprint, a grant or\n a pitch, and to checking a paper or statistic before relying on it.\n\n Runs no model, costs nothing, and returns immediately. The queries worked out from\n its output are what zetesis_evidence takes as input. With as_of set, the apparatus\n and rubric are framed for judging the claim as it stood at the end of that year.\n\n Args:\n claim: the claim to evaluate, in one or two sentences.\n context: optional background (stage, field, the decision at hand).\n as_of: optional year. Set it to evaluate the claim as it stood at the end of that\n year, grading how it was derived at the time rather than how it turned out.\n claim_class: optional override for the routed claim class. The response lists\n every class; pass one back if the routing picked wrongly, or if nothing\n routed.",
"inputSchema": {
"properties": {
"as_of": {
"default": 0,
"description": "Year the claim was made. Set it for any claim that is not brand new: it restricts evidence to what existed then, which keeps later outcomes out of the reading and sharply improves what is retrieved.",
"title": "As Of",
"type": "integer"
},
"claim": {
"description": "The claim to evaluate, in one or two sentences.",
"title": "Claim",
"type": "string"
},
"claim_class": {
"default": "",
"description": "Optional override for the routed claim class. The response lists every class; pass one back if the routing picked wrongly.",
"title": "Claim Class",
"type": "string"
},
"context": {
"default": "",
"description": "Optional background: stage, field, or the decision at hand.",
"title": "Context",
"type": "string"
}
},
"required": [
"claim"
],
"title": "zetesis_scopeArguments",
"type": "object"
},
"name": "zetesis_scope",
"outputSchema": {
"properties": {
"result": {
"title": "Result",
"type": "string"
}
},
"required": [
"result"
],
"title": "zetesis_scopeOutput",
"type": "object"
}
}
]
}Verify it yourself
curl -s https://api.teppi.xyz/v1/evidence/sha256:ac8613d6c318b0922a4641193bd4b5e0653fad953e704eb0798f9fb3af8c2989 | sha256sum