Blossom Research

Find patent families

search_patents

Read-onlyIdempotent

Search Blossom's patent families on psychedelic medicine (psilocybin, MDMA, LSD, DMT, ketamine for mental health and pain, ibogaine and the new molecules built on them). One record is one family: one invention in all countries. Use to LIST families by compound, compound class, indication, owner, inventor, patent type, office, stage or priority date. Use publicationNumbers for an exact patent or application number. Each row has the lead publication, the stage (granted or application), the owners, the priority date, the estimated end of the 20-year term and a short Blossom summary. computedStatus is a computation from the publications and the term, not a check of an office register: report it as computed. legalStatus is set only when a register or an editor gave it. For 'how many?' questions use count_patents instead.

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "search_patents"
  }
}

Input Schema

Name	Required	Description	Default
`term`	No	"ended" = the estimated 20-year term is over; "running" = it is not, or there is no estimate. The estimate is the filing date plus 20 years: adjustments, extensions and unpaid fees change the real date.
`limit`	No
`owner`	No	Owner of the family: an organisation name or a Blossom organisation slug ("COMPASS Pathways", "MindMed", "gh-research", "Sandoz"). A name, other name or slug of a Blossom organisation record matches the families that are linked to it. Any other name matches the owner names as printed on the publications, word by word.
`query`	No	Text search across the title, the Blossom summary, the abstract, the first claim, owner and inventor names, compounds, indications and publication numbers. For an exact publication number use publicationNumbers.
`stage`	No	"granted" = at least one office has granted a patent of the family; "application" = applications only. This is the reliable stage from the document kinds, not a check that the patent is in force.
`topics`	No	Topic slugs of the indications that the claims or the stated purpose name (e.g. "depressive-disorders", "ptsd"). Under half of the families name an indication.
`offices`	No	Offices where the family has a publication, as two-letter codes ("US", "EP", "WO" for a PCT application, "JP", "CN"). Country names are auto-mapped.
`orderBy`	No	Sort order. "relevance" = best text match (newest priority date first when there is no query). "newest" / "oldest" = by priority date. "mostCited" = families that later patents cite most.	relevance
`inventor`	No	Inventor name ("Albert Hofmann", "Nichols"). A name that identifies one Blossom person record matches the families linked to it; any other name matches the inventor names as printed, word by word.
`compounds`	No	Compound slugs that the claims are directed at (e.g. "psilocybin", "mdma", "esketamine"). Common names are auto-mapped. A family whose claims are for a new molecule often has a compound class and no compound: use compoundClasses for those.
`priorityTo`	No	Inclusive upper bound of the priority date (YYYY-MM-DD).
`patentTypes`	No	What the claims cover: "derivative" (a new compound, salt, crystal form or prodrug), "delivery-formulation", "method-of-treatment", "method-of-manufacture", "drug-combination", "other".
`priorityFrom`	No	Inclusive lower bound of the priority date (YYYY-MM-DD). The priority date is the date that orders inventions; a family with no priority date in the source is not in a dated result.
`grantedOffices`	No	Offices that granted a patent of the family ("US", "EP", ...).
`compoundClasses`	No	Chemical family of what is claimed: "tryptamine" (psilocybin, DMT, 5-MeO-DMT and analogues), "lysergamide" (LSD type), "phenethylamine" (mescaline, 2C family), "entactogen" (MDMA type), "arylcyclohexylamine" (ketamine family), "iboga-type", "other". "general" = claims that are about no single compound (any psychedelic, a tool or a device).
`publicationNumbers`	No	Publication or patent numbers, in any usual form ("US 10,519,175 B2", "US10519175B2", "WO2019073379A1", "EP3927338"). Each number matches the family that has it as a member. A number with no kind code matches every kind.

Output Schema

Name	Required	Description	Default
`result`	Yes

Schema Changelog

Changes observed during successful MCP inspections.

First observedNov 7, 2026

TDQS

A4.6/5.0

Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Well beyond what readOnlyHint/idempotentHint convey: it explains that computedStatus is derived from publications and term rather than an office register and must be reported as computed, that legalStatus is populated only when a register or editor supplied it, and that topics cover 'under half of the families.' These are real caveats an agent would otherwise misreport. It does not cover pagination or result-size behavior, keeping it short of a 5.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loaded with identity and the one-record-one-family model, then facets, then caveats, then the count_patents redirect. Dense but nearly every clause earns its place; the long parenthetical compound list in the opening sentence is the one mildly expendable element.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 16-parameter search with zero required params and a 94%-documented schema, the description covers the record model, the facet surface, the computed-vs-registered status distinction, and the counting alternative. An agent can call it correctly from this alone, though aggregate_patents as a sibling path is never addressed.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 94%, so the baseline is 3, but the description adds genuine semantics on top: it clarifies that publicationNumbers is the exact-number path while query is fuzzy text search, and it describes what each row returns (lead publication, stage, owners, priority date, estimated term end, summary). That is meaning beyond the schema rather than a restatement of it.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description opens with a specific verb and resource ('Search Blossom's patent families'), scopes the corpus (psychedelic medicine compounds), and states a crucial modeling fact: 'One record is one family: one invention in all countries.' That distinguishes it from get_patent and from the aggregate/count siblings without needing the schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It enumerates the filterable facets ('LIST families by compound, compound class, indication, owner, inventor, patent type, office, stage or priority date'), routes exact-number lookups to publicationNumbers, and explicitly redirects counting questions: 'For how many? questions use count_patents instead.' Only aggregate_patents goes unmentioned, which is a small omission against otherwise explicit routing.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All Blossom Research tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources

Need Help?