modulector-sdk 2.3.0__tar.gz → 2.5.2__tar.gz

{modulector_sdk-2.3.0/src/modulector_sdk.egg-info → modulector_sdk-2.5.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: modulector-sdk
-Version: 2.3.0
+Version: 2.5.2
 Summary: Typed Python SDK for the Modulector API.
 Author: omicsdatascience
 License-Expression: MIT
@@ -25,6 +25,7 @@ Classifier: Typing :: Typed
 Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE
+Requires-Dist: mcp==1.27.2
 Requires-Dist: requests==2.31.0
 Dynamic: license-file
 
@@ -63,6 +64,46 @@ MODULECTOR_API_BASE_URL=https://your-modulector.example.org python script.py
 Every service function also accepts a `base_url` keyword argument for
 per-request overrides.
 
+## MCP server
+
+The SDK also installs a Model Context Protocol server that exposes the same
+service functions as LLM-callable tools.
+
+For local stdio MCP clients:
+
+```bash
+modulector-mcp
+```
+
+Example client configuration:
+
+```json
+{
+  "mcpServers": {
+    "modulector": {
+      "command": "modulector-mcp"
+    }
+  }
+}
+```
+
+For Streamable HTTP clients:
+
+```bash
+modulector-mcp --transport streamable-http --host 127.0.0.1 --port 8000
+```
+
+Then connect the client to `http://127.0.0.1:8000/mcp`.
+
+The server includes tools for miRNA target interactions, miRNA details and
+aliases, miRNA identifier resolution, methylation site lookup and annotation,
+disease associations, drug associations, and PubMed news subscriptions. Use
+`MODULECTOR_API_BASE_URL` or each tool's `base_url` argument to target a custom
+Modulector deployment.
+
+When deployed with the root Docker Compose stack, nginx proxies the MCP
+Streamable HTTP endpoint at `https://<mydomain.com>/mcp`.
+
 ## Development
 
 Build this SDK from the `sdk/` directory:

modulector_sdk-2.5.2/README.md

@@ -0,0 +1,84 @@
+# Modulector SDK
+
+Typed Python SDK for the Modulector API. The package contains client helpers for
+querying miRNA target interactions, miRNA aliases, methylation sites, disease
+associations, drug associations, and PubMed subscriptions without installing the
+Django backend dependencies.
+
+## Installation
+
+```bash
+pip install modulector-sdk
+```
+
+## Usage
+
+```python
+from modulector_sdk import get_mirna_details, get_mirna_target_interactions
+
+details = get_mirna_details("hsa-miR-21-5p")
+interactions = get_mirna_target_interactions(
+    mirna="hsa-miR-21-5p",
+    gene="PTEN",
+    include_pubmeds=True,
+)
+```
+
+Set `MODULECTOR_API_BASE_URL` to target a different Modulector deployment:
+
+```bash
+MODULECTOR_API_BASE_URL=https://your-modulector.example.org python script.py
+```
+
+Every service function also accepts a `base_url` keyword argument for
+per-request overrides.
+
+## MCP server
+
+The SDK also installs a Model Context Protocol server that exposes the same
+service functions as LLM-callable tools.
+
+For local stdio MCP clients:
+
+```bash
+modulector-mcp
+```
+
+Example client configuration:
+
+```json
+{
+  "mcpServers": {
+    "modulector": {
+      "command": "modulector-mcp"
+    }
+  }
+}
+```
+
+For Streamable HTTP clients:
+
+```bash
+modulector-mcp --transport streamable-http --host 127.0.0.1 --port 8000
+```
+
+Then connect the client to `http://127.0.0.1:8000/mcp`.
+
+The server includes tools for miRNA target interactions, miRNA details and
+aliases, miRNA identifier resolution, methylation site lookup and annotation,
+disease associations, drug associations, and PubMed news subscriptions. Use
+`MODULECTOR_API_BASE_URL` or each tool's `base_url` argument to target a custom
+Modulector deployment.
+
+When deployed with the root Docker Compose stack, nginx proxies the MCP
+Streamable HTTP endpoint at `https://<mydomain.com>/mcp`.
+
+## Development
+
+Build this SDK from the `sdk/` directory:
+
+```bash
+uv build
+```
+
+The generated source distribution and wheel are written to `sdk/dist/`.

{modulector_sdk-2.3.0 → modulector_sdk-2.5.2}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "modulector-sdk"
-version = "2.3.0"
+version = "2.5.2"
 description = "Typed Python SDK for the Modulector API."
 readme = "README.md"
 requires-python = ">=3.10"
@@ -36,6 +36,7 @@ classifiers = [
     "Typing :: Typed",
 ]
 dependencies = [
+    "mcp==1.27.2",
     "requests==2.31.0",
 ]
 
@@ -45,6 +46,9 @@ Homepage = "https://modulector.multiomix.org/"
 Issues = "https://github.com/omics-datascience/modulector/issues"
 Repository = "https://github.com/omics-datascience/modulector"
 
+[project.scripts]
+modulector-mcp = "modulector_sdk.mcp_server:main"
+
 [tool.setuptools.packages.find]
 where = ["src"]
 

modulector_sdk-2.5.2/src/modulector_sdk/mcp_server.py

@@ -0,0 +1,538 @@
+"""MCP server exposing Modulector SDK services to LLM clients."""
+
+from __future__ import annotations
+
+import argparse
+import os
+from typing import Any, Final, Literal, TypeVar, cast
+from urllib.parse import urlparse
+
+from mcp.server.fastmcp import FastMCP
+from mcp.server.transport_security import TransportSecuritySettings
+
+from . import services
+from .utils import PaginatedResponse
+
+T = TypeVar("T")
+Transport = Literal["stdio", "sse", "streamable-http"]
+
+SERVER_INSTRUCTIONS: Final[str] = (
+    "Use Modulector to help researchers inspect human miRNA target "
+    "interactions, miRNA aliases, disease and drug associations, and "
+    "Illumina methylation site annotations. Prefer finder tools when the "
+    "user gives partial or uncertain identifiers, then use the detail or "
+    "paginated tools with explicit identifiers. The returned data comes from "
+    "the configured Modulector API deployment."
+)
+PUBLIC_BASE_URL: Final[str] = os.getenv(
+    "MODULECTOR_PUBLIC_BASE_URL",
+    services.MODULECTOR_API_BASE_URL,
+)
+
+mcp = FastMCP(
+    name="modulector",
+    instructions=SERVER_INSTRUCTIONS,
+    website_url=PUBLIC_BASE_URL,
+)
+
+
+@mcp.resource(
+    "modulector://about",
+    title="About Modulector",
+    description="Overview of the Modulector data exposed by this MCP server.",
+    mime_type="text/markdown",
+)
+def about_modulector() -> str:
+    """Return a concise overview of this MCP server.
+
+    :return: Markdown overview for LLM context.
+    """
+
+    return (
+        "# Modulector\n\n"
+        "Modulector provides programmatic access to curated human miRNA target "
+        "interactions, miRNA aliases, miRNA-disease associations, "
+        "miRNA-drug associations, and Illumina Infinium MethylationEPIC 2.0 "
+        "site annotations.\n\n"
+        f"Default API base URL: `{services.MODULECTOR_API_BASE_URL}`\n\n"
+        f"Public website URL: `{PUBLIC_BASE_URL}`\n\n"
+        "Use the finder tools for partial identifiers, then query detail or "
+        "paginated tools with exact miRNA, gene, disease, drug, or methylation "
+        "site filters."
+    )
+
+
+@mcp.prompt(
+    name="modulector_research_plan",
+    title="Modulector Research Plan",
+    description="Guide an LLM through a focused Modulector research lookup.",
+)
+def modulector_research_plan(research_question: str) -> str:
+    """Return a prompt for planning a Modulector lookup.
+
+    :param research_question: Research question to investigate.
+    :return: Prompt text that guides tool selection and result synthesis.
+    """
+
+    return (
+        "Plan a Modulector lookup for this research question:\n\n"
+        f"{research_question}\n\n"
+        "First identify whether the task involves miRNA identifiers, gene "
+        "symbols, diseases, drugs, or methylation sites. Use finder or alias "
+        "tools to normalize uncertain identifiers. Then retrieve the smallest "
+        "useful page of records, include PubMed links only when needed, and "
+        "summarize the evidence with clear caveats about database scope."
+    )
+
+
+@mcp.tool()
+def get_mirna_target_interactions(
+    mirna: str | None = None,
+    gene: str | None = None,
+    score: float | None = None,
+    include_pubmeds: bool | None = None,
+    ordering: str | None = None,
+    search: str | None = None,
+    page: int | None = None,
+    page_size: int | None = None,
+    base_url: str | None = None,
+    timeout: float = 30.0,
+) -> dict[str, Any]:
+    """Return miRNA-gene target interactions from Modulector.
+
+    :param mirna: miRNA accession ID or miRBase mature name.
+    :param gene: Gene symbol.
+    :param score: Minimum MirDIP score between 0 and 1.
+    :param include_pubmeds: Whether to include PubMed URLs for interactions.
+    :param ordering: Comma-separated ordering fields, such as ``-score,gene``.
+    :param search: Search term for supported server-side fields.
+    :param page: Page number to request.
+    :param page_size: Number of records to request.
+    :param base_url: Optional Modulector API deployment URL.
+    :param timeout: Request timeout in seconds.
+    :return: Paginated interaction records.
+    """
+
+    return _page_to_dict(
+        services.get_mirna_target_interactions(
+            base_url=_base_url(base_url),
+            mirna=mirna,
+            gene=gene,
+            score=score,
+            include_pubmeds=include_pubmeds,
+            ordering=ordering,
+            search=search,
+            page=page,
+            page_size=page_size,
+            timeout=timeout,
+        )
+    )
+
+
+@mcp.tool()
+def get_mirna_details(
+    mirna: str,
+    base_url: str | None = None,
+    timeout: float = 30.0,
+) -> dict[str, Any]:
+    """Return sequence, aliases, accession, and source links for one miRNA.
+
+    :param mirna: miRNA identifier, such as a mature miRNA ID or accession ID.
+    :param base_url: Optional Modulector API deployment URL.
+    :param timeout: Request timeout in seconds.
+    :return: miRNA detail record.
+    """
+
+    return cast(
+        dict[str, Any],
+        services.get_mirna_details(
+            mirna,
+            base_url=_base_url(base_url),
+            timeout=timeout,
+        ),
+    )
+
+
+@mcp.tool()
+def get_mirna_aliases(
+    mature_mirna: str | None = None,
+    mirbase_accession_id: str | None = None,
+    previous_mature_mirna: str | None = None,
+    search: str | None = None,
+    ordering: str | None = None,
+    page: int | None = None,
+    page_size: int | None = None,
+    base_url: str | None = None,
+    timeout: float = 30.0,
+) -> dict[str, Any]:
+    """Return miRBase accession and mature miRNA alias records.
+
+    :param mature_mirna: Exact mature miRNA filter.
+    :param mirbase_accession_id: Exact miRBase accession ID filter.
+    :param previous_mature_mirna: Exact previous mature miRNA filter.
+    :param search: Search across mature IDs, previous IDs, and accessions.
+    :param ordering: Comma-separated ordering fields.
+    :param page: Page number to request.
+    :param page_size: Number of records to request.
+    :param base_url: Optional Modulector API deployment URL.
+    :param timeout: Request timeout in seconds.
+    :return: Paginated miRNA alias records.
+    """
+
+    return _page_to_dict(
+        services.get_mirna_aliases(
+            base_url=_base_url(base_url),
+            mature_mirna=mature_mirna,
+            mirbase_accession_id=mirbase_accession_id,
+            previous_mature_mirna=previous_mature_mirna,
+            search=search,
+            ordering=ordering,
+            page=page,
+            page_size=page_size,
+            timeout=timeout,
+        )
+    )
+
+
+@mcp.tool()
+def find_mirna_codes(
+    query: str,
+    limit: int | None = None,
+    base_url: str | None = None,
+    timeout: float = 30.0,
+) -> list[str]:
+    """Return miRNA identifiers matching a partial search string.
+
+    :param query: miRNA search string.
+    :param limit: Maximum number of identifiers to return.
+    :param base_url: Optional Modulector API deployment URL.
+    :param timeout: Request timeout in seconds.
+    :return: Matching miRNA IDs or accession IDs.
+    """
+
+    return services.find_mirna_codes(
+        query,
+        base_url=_base_url(base_url),
+        limit=limit,
+        timeout=timeout,
+    )
+
+
+@mcp.tool()
+def get_mirna_codes(
+    mirna_codes: list[str],
+    base_url: str | None = None,
+    timeout: float = 30.0,
+) -> dict[str, str | None]:
+    """Resolve miRNA identifiers to approved miRBase accessions.
+
+    :param mirna_codes: miRNA identifiers to resolve.
+    :param base_url: Optional Modulector API deployment URL.
+    :param timeout: Request timeout in seconds.
+    :return: Mapping from requested identifiers to accession IDs or null.
+    """
+
+    return services.get_mirna_codes(
+        mirna_codes,
+        base_url=_base_url(base_url),
+        timeout=timeout,
+    )
+
+
+@mcp.tool()
+def find_methylation_sites(
+    query: str,
+    limit: int | None = None,
+    base_url: str | None = None,
+    timeout: float = 30.0,
+) -> list[str]:
+    """Return methylation site names matching a partial search string.
+
+    :param query: Methylation site search string.
+    :param limit: Maximum number of site names to return.
+    :param base_url: Optional Modulector API deployment URL.
+    :param timeout: Request timeout in seconds.
+    :return: Matching methylation site names.
+    """
+
+    return services.find_methylation_sites(
+        query,
+        base_url=_base_url(base_url),
+        limit=limit,
+        timeout=timeout,
+    )
+
+
+@mcp.tool()
+def get_methylation_sites(
+    methylation_sites: list[str],
+    base_url: str | None = None,
+    timeout: float = 30.0,
+) -> dict[str, list[str]]:
+    """Resolve methylation identifiers to EPIC 2.0 site names.
+
+    :param methylation_sites: Illumina methylation site names or identifiers.
+    :param base_url: Optional Modulector API deployment URL.
+    :param timeout: Request timeout in seconds.
+    :return: Mapping from requested identifiers to current EPIC 2.0 names.
+    """
+
+    return services.get_methylation_sites(
+        methylation_sites,
+        base_url=_base_url(base_url),
+        timeout=timeout,
+    )
+
+
+@mcp.tool()
+def get_methylation_site_genes(
+    methylation_sites: list[str],
+    base_url: str | None = None,
+    timeout: float = 30.0,
+) -> dict[str, list[str]]:
+    """Return genes associated with methylation site identifiers.
+
+    :param methylation_sites: Illumina methylation site names or identifiers.
+    :param base_url: Optional Modulector API deployment URL.
+    :param timeout: Request timeout in seconds.
+    :return: Mapping from requested site identifiers to associated genes.
+    """
+
+    return services.get_methylation_site_genes(
+        methylation_sites,
+        base_url=_base_url(base_url),
+        timeout=timeout,
+    )
+
+
+@mcp.tool()
+def get_methylation_details(
+    methylation_site: str,
+    base_url: str | None = None,
+    timeout: float = 30.0,
+) -> dict[str, Any]:
+    """Return detailed annotations for one methylation site.
+
+    :param methylation_site: Methylation site name from EPIC 2.0.
+    :param base_url: Optional Modulector API deployment URL.
+    :param timeout: Request timeout in seconds.
+    :return: Methylation site detail record.
+    """
+
+    return cast(
+        dict[str, Any],
+        services.get_methylation_details(
+            methylation_site,
+            base_url=_base_url(base_url),
+            timeout=timeout,
+        ),
+    )
+
+
+@mcp.tool()
+def get_diseases(
+    mirna: str | None = None,
+    search: str | None = None,
+    ordering: str | None = None,
+    page: int | None = None,
+    page_size: int | None = None,
+    base_url: str | None = None,
+    timeout: float = 30.0,
+) -> dict[str, Any]:
+    """Return miRNA and human disease association records.
+
+    :param mirna: miRNA code or accession ID filter.
+    :param search: Case-insensitive disease search term.
+    :param ordering: Comma-separated ordering fields.
+    :param page: Page number to request.
+    :param page_size: Number of records to request.
+    :param base_url: Optional Modulector API deployment URL.
+    :param timeout: Request timeout in seconds.
+    :return: Paginated disease association records.
+    """
+
+    return _page_to_dict(
+        services.get_diseases(
+            base_url=_base_url(base_url),
+            mirna=mirna,
+            search=search,
+            ordering=ordering,
+            page=page,
+            page_size=page_size,
+            timeout=timeout,
+        )
+    )
+
+
+@mcp.tool()
+def get_drugs(
+    mirna: str | None = None,
+    fda_approved: bool | None = None,
+    search: str | None = None,
+    ordering: str | None = None,
+    page: int | None = None,
+    page_size: int | None = None,
+    base_url: str | None = None,
+    timeout: float = 30.0,
+) -> dict[str, Any]:
+    """Return drug or small molecule records affecting miRNA expression.
+
+    :param mirna: miRNA code or accession ID filter.
+    :param fda_approved: Optional FDA approval filter.
+    :param search: Search term for condition, molecule, or expression fields.
+    :param ordering: Comma-separated ordering fields.
+    :param page: Page number to request.
+    :param page_size: Number of records to request.
+    :param base_url: Optional Modulector API deployment URL.
+    :param timeout: Request timeout in seconds.
+    :return: Paginated drug association records.
+    """
+
+    return _page_to_dict(
+        services.get_drugs(
+            base_url=_base_url(base_url),
+            mirna=mirna,
+            fda_approved=fda_approved,
+            search=search,
+            ordering=ordering,
+            page=page,
+            page_size=page_size,
+            timeout=timeout,
+        )
+    )
+
+
+def main() -> None:
+    """Run the Modulector MCP server.
+
+    :return: Nothing.
+    """
+
+    args = _parse_args()
+    transport = cast(Transport, args.transport)
+    if transport != "stdio":
+        mcp.settings.host = args.host
+        mcp.settings.port = args.port
+        mcp.settings.json_response = args.json_response
+        _configure_http_transport_security()
+    mcp.run(transport=transport)
+
+
+def _parse_args() -> argparse.Namespace:
+    """Parse command-line arguments for the MCP server.
+
+    :return: Parsed command-line arguments.
+    """
+
+    parser = argparse.ArgumentParser(
+        description="Run the Modulector MCP server.",
+    )
+    parser.add_argument(
+        "--transport",
+        choices=("stdio", "sse", "streamable-http"),
+        default="stdio",
+        help="MCP transport to use. Defaults to stdio.",
+    )
+    parser.add_argument(
+        "--host",
+        default="127.0.0.1",
+        help="Host for HTTP transports. Defaults to 127.0.0.1.",
+    )
+    parser.add_argument(
+        "--port",
+        type=int,
+        default=8000,
+        help="Port for HTTP transports. Defaults to 8000.",
+    )
+    parser.add_argument(
+        "--json-response",
+        action="store_true",
+        help="Use JSON responses for Streamable HTTP instead of SSE streams.",
+    )
+    return parser.parse_args()
+
+
+def _base_url(base_url: str | None) -> str:
+    """Return the requested or default Modulector API base URL.
+
+    :param base_url: Optional Modulector API deployment URL.
+    :return: API base URL.
+    """
+
+    return base_url or services.MODULECTOR_API_BASE_URL
+
+
+def _configure_http_transport_security() -> None:
+    """Configure DNS rebinding protection for HTTP MCP transports.
+
+    :return: None.
+    """
+
+    mcp.settings.transport_security = TransportSecuritySettings(
+        allowed_hosts=_allowed_hosts(),
+        allowed_origins=_allowed_origins(),
+    )
+
+
+def _allowed_hosts() -> list[str]:
+    """Return allowed Host headers for the MCP HTTP transport.
+
+    :return: Host header allow-list.
+    """
+
+    configured_hosts = _csv_env("MODULECTOR_MCP_ALLOWED_HOSTS")
+    if configured_hosts:
+        return configured_hosts
+
+    parsed_public_url = urlparse(PUBLIC_BASE_URL)
+    public_host = parsed_public_url.netloc
+    hosts = ["127.0.0.1:*", "localhost:*", "[::1]:*"]
+    if public_host:
+        hosts.append(public_host)
+    return hosts
+
+
+def _allowed_origins() -> list[str]:
+    """Return allowed Origin headers for the MCP HTTP transport.
+
+    :return: Origin header allow-list.
+    """
+
+    configured_origins = _csv_env("MODULECTOR_MCP_ALLOWED_ORIGINS")
+    if configured_origins:
+        return configured_origins
+
+    parsed_public_url = urlparse(PUBLIC_BASE_URL)
+    origins = ["http://127.0.0.1:*", "http://localhost:*", "http://[::1]:*"]
+    if parsed_public_url.scheme and parsed_public_url.netloc:
+        origins.append(f"{parsed_public_url.scheme}://{parsed_public_url.netloc}")
+    return origins
+
+
+def _csv_env(name: str) -> list[str]:
+    """Return comma-separated environment variable values.
+
+    :param name: Environment variable name.
+    :return: Non-empty stripped values.
+    """
+
+    return [value.strip() for value in os.getenv(name, "").split(",") if value.strip()]
+
+
+def _page_to_dict(page: PaginatedResponse[T]) -> dict[str, Any]:
+    """Convert a paginated SDK response into a JSON-serializable mapping.
+
+    :param page: SDK paginated response.
+    :return: Mapping with pagination metadata and results.
+    """
+
+    return {
+        "count": page.count,
+        "next": page.next,
+        "previous": page.previous,
+        "results": page.results,
+    }
+
+
+if __name__ == "__main__":
+    main()

{modulector_sdk-2.3.0 → modulector_sdk-2.5.2}/src/modulector_sdk/utils.py

@@ -191,6 +191,15 @@ def _with_pagination_params(
     page: int | None,
     page_size: int | None,
 ) -> dict[str, Any] | None:
+    """
+    Return a new params dict with pagination parameters added, or None if no params.
+    :param params: Parameters to include in the request, or None for no parameters.
+    :param page: Page number to include in the request, or None for no page parameter.
+    :param page_size: Number of items per page to include in the request, or None for no page_size parameter.
+    :raises ValueError: If page is less than 1.
+    :raises ValueError: If page_size is not between 1 and 1000.
+    :return: A new params dict with pagination parameters added, or None if no params.
+    """
     if page is not None and page < 1:
         raise ValueError("page must be greater than 0")
     if page_size is not None and not 1 <= page_size <= 1000:
@@ -225,9 +234,7 @@ def _parse_paginated_response(payload: JSON) -> PaginatedResponse[Any]:
     if next_url is not None and not isinstance(next_url, str):
         raise ValueError("paginated response field 'next' must be a string or null")
     if previous_url is not None and not isinstance(previous_url, str):
-        raise ValueError(
-            "paginated response field 'previous' must be a string or null"
-        )
+        raise ValueError("paginated response field 'previous' must be a string or null")
 
     return PaginatedResponse(
         count=int(payload["count"]),

{modulector_sdk-2.3.0 → modulector_sdk-2.5.2/src/modulector_sdk.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: modulector-sdk
-Version: 2.3.0
+Version: 2.5.2
 Summary: Typed Python SDK for the Modulector API.
 Author: omicsdatascience
 License-Expression: MIT
@@ -25,6 +25,7 @@ Classifier: Typing :: Typed
 Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE
+Requires-Dist: mcp==1.27.2
 Requires-Dist: requests==2.31.0
 Dynamic: license-file
 
@@ -63,6 +64,46 @@ MODULECTOR_API_BASE_URL=https://your-modulector.example.org python script.py
 Every service function also accepts a `base_url` keyword argument for
 per-request overrides.
 
+## MCP server
+
+The SDK also installs a Model Context Protocol server that exposes the same
+service functions as LLM-callable tools.
+
+For local stdio MCP clients:
+
+```bash
+modulector-mcp
+```
+
+Example client configuration:
+
+```json
+{
+  "mcpServers": {
+    "modulector": {
+      "command": "modulector-mcp"
+    }
+  }
+}
+```
+
+For Streamable HTTP clients:
+
+```bash
+modulector-mcp --transport streamable-http --host 127.0.0.1 --port 8000
+```
+
+Then connect the client to `http://127.0.0.1:8000/mcp`.
+
+The server includes tools for miRNA target interactions, miRNA details and
+aliases, miRNA identifier resolution, methylation site lookup and annotation,
+disease associations, drug associations, and PubMed news subscriptions. Use
+`MODULECTOR_API_BASE_URL` or each tool's `base_url` argument to target a custom
+Modulector deployment.
+
+When deployed with the root Docker Compose stack, nginx proxies the MCP
+Streamable HTTP endpoint at `https://<mydomain.com>/mcp`.
+
 ## Development
 
 Build this SDK from the `sdk/` directory:

{modulector_sdk-2.3.0 → modulector_sdk-2.5.2}/src/modulector_sdk.egg-info/SOURCES.txt

@@ -2,11 +2,13 @@ LICENSE
 README.md
 pyproject.toml
 src/modulector_sdk/__init__.py
+src/modulector_sdk/mcp_server.py
 src/modulector_sdk/py.typed
 src/modulector_sdk/services.py
 src/modulector_sdk/utils.py
 src/modulector_sdk.egg-info/PKG-INFO
 src/modulector_sdk.egg-info/SOURCES.txt
 src/modulector_sdk.egg-info/dependency_links.txt
+src/modulector_sdk.egg-info/entry_points.txt
 src/modulector_sdk.egg-info/requires.txt
 src/modulector_sdk.egg-info/top_level.txt

modulector_sdk-2.5.2/src/modulector_sdk.egg-info/entry_points.txt

@@ -0,0 +1,2 @@
+[console_scripts]
+modulector-mcp = modulector_sdk.mcp_server:main

{modulector_sdk-2.3.0 → modulector_sdk-2.5.2}/src/modulector_sdk.egg-info/requires.txt

@@ -1 +1,2 @@
+mcp==1.27.2
 requests==2.31.0

modulector_sdk-2.3.0/README.md

@@ -1,44 +0,0 @@
-# Modulector SDK
-
-Typed Python SDK for the Modulector API. The package contains client helpers for
-querying miRNA target interactions, miRNA aliases, methylation sites, disease
-associations, drug associations, and PubMed subscriptions without installing the
-Django backend dependencies.
-
-## Installation
-
-```bash
-pip install modulector-sdk
-```
-
-## Usage
-
-```python
-from modulector_sdk import get_mirna_details, get_mirna_target_interactions
-
-details = get_mirna_details("hsa-miR-21-5p")
-interactions = get_mirna_target_interactions(
-    mirna="hsa-miR-21-5p",
-    gene="PTEN",
-    include_pubmeds=True,
-)
-```
-
-Set `MODULECTOR_API_BASE_URL` to target a different Modulector deployment:
-
-```bash
-MODULECTOR_API_BASE_URL=https://your-modulector.example.org python script.py
-```
-
-Every service function also accepts a `base_url` keyword argument for
-per-request overrides.
-
-## Development
-
-Build this SDK from the `sdk/` directory:
-
-```bash
-uv build
-```
-
-The generated source distribution and wheel are written to `sdk/dist/`.