academia-mcp 1.11.4__tar.gz → 1.11.5__tar.gz

{academia_mcp-1.11.4 → academia_mcp-1.11.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: academia-mcp
-Version: 1.11.4
+Version: 1.11.5
 Summary: MCP server that provides different tools to search for scientific publications
 Author-email: Ilya Gusev <phoenixilya@gmail.com>
 Project-URL: Homepage, https://github.com/IlyaGusev/academia_mcp

{academia_mcp-1.11.4 → academia_mcp-1.11.5}/academia_mcp/server.py

@@ -17,6 +17,7 @@ from academia_mcp.tools.s2 import (
     s2_get_references,
     s2_corpus_id_from_arxiv_id,
     s2_get_info,
+    s2_search,
 )
 from academia_mcp.tools.hf_datasets_search import hf_datasets_search
 from academia_mcp.tools.anthology_search import anthology_search
@@ -86,6 +87,7 @@ def create_server(
     server.add_tool(s2_get_citations, structured_output=True)
     server.add_tool(s2_get_references, structured_output=True)
     server.add_tool(s2_get_info, structured_output=True)
+    server.add_tool(s2_search, structured_output=True)
     server.add_tool(s2_corpus_id_from_arxiv_id)
     server.add_tool(hf_datasets_search)
     server.add_tool(anthology_search)

{academia_mcp-1.11.4 → academia_mcp-1.11.5}/academia_mcp/tools/__init__.py

@@ -2,7 +2,13 @@ from .arxiv_search import arxiv_search
 from .anthology_search import anthology_search
 from .arxiv_download import arxiv_download
 from .hf_datasets_search import hf_datasets_search
-from .s2 import s2_get_references, s2_get_citations, s2_corpus_id_from_arxiv_id, s2_get_info
+from .s2 import (
+    s2_get_references,
+    s2_get_citations,
+    s2_corpus_id_from_arxiv_id,
+    s2_get_info,
+    s2_search,
+)
 from .document_qa import document_qa
 from .latex import (
     compile_latex,
@@ -26,6 +32,7 @@ __all__ = [
     "s2_get_citations",
     "s2_corpus_id_from_arxiv_id",
     "s2_get_info",
+    "s2_search",
     "hf_datasets_search",
     "document_qa",
     "compile_latex",

{academia_mcp-1.11.4 → academia_mcp-1.11.5}/academia_mcp/tools/s2.py

@@ -8,10 +8,12 @@ from pydantic import BaseModel, Field
 from academia_mcp.utils import get_with_retries
 
 
-PAPER_URL_TEMPLATE = "https://api.semanticscholar.org/graph/v1/paper/{paper_id}?fields={fields}"
-CITATIONS_URL_TEMPLATE = "https://api.semanticscholar.org/graph/v1/paper/{paper_id}/citations?fields={fields}&offset={offset}&limit={limit}"
-REFERENCES_URL_TEMPLATE = "https://api.semanticscholar.org/graph/v1/paper/{paper_id}/references?fields={fields}&offset={offset}&limit={limit}"
-FIELDS = "title,authors,externalIds,venue,citationCount,publicationDate"
+BASE_URL = "https://api.semanticscholar.org/graph/v1"
+PAPER_URL_TEMPLATE = "{base_url}/paper/{paper_id}"
+CITATIONS_URL_TEMPLATE = "{base_url}/paper/{paper_id}/citations"
+REFERENCES_URL_TEMPLATE = "{base_url}/paper/{paper_id}/references"
+SEARCH_URL_TEMPLATE = "{base_url}/paper/search"
+FIELDS = "paperId,title,authors,externalIds,venue,citationCount,publicationDate"
 
 
 class S2PaperInfo(BaseModel):  # type: ignore
@@ -38,7 +40,10 @@ def _format_authors(authors: List[Dict[str, Any]]) -> List[str]:
 
 
 def _clean_entry(entry: Dict[str, Any]) -> S2PaperInfo:
-    entry = entry["citingPaper"] if "citingPaper" in entry else entry["citedPaper"]
+    if "citingPaper" in entry:
+        entry = entry["citingPaper"]
+    elif "citedPaper" in entry:
+        entry = entry["citedPaper"]
     external_ids = entry.get("externalIds")
     if not external_ids:
         external_ids = dict()
@@ -88,17 +93,17 @@ def s2_get_citations(
         arxiv_id = arxiv_id.split("v")[0]
     paper_id = f"arxiv:{arxiv_id}"
 
-    url = CITATIONS_URL_TEMPLATE.format(
-        paper_id=paper_id, fields=FIELDS, offset=offset, limit=limit
-    )
-    response = get_with_retries(url)
+    url = CITATIONS_URL_TEMPLATE.format(base_url=BASE_URL, paper_id=paper_id)
+    payload = {"fields": FIELDS, "offset": offset, "limit": limit}
+    response = get_with_retries(url, params=payload)
     result = response.json()
     entries = result["data"]
     total_count = len(result["data"]) + result["offset"]
 
     if "next" in result:
-        paper_url = PAPER_URL_TEMPLATE.format(paper_id=paper_id, fields=FIELDS)
-        paper_response = get_with_retries(paper_url)
+        paper_url = PAPER_URL_TEMPLATE.format(base_url=BASE_URL, paper_id=paper_id)
+        payload = {"fields": FIELDS}
+        paper_response = get_with_retries(paper_url, params=payload)
         paper_result = paper_response.json()
         total_count = paper_result["citationCount"]
 
@@ -123,10 +128,9 @@ def s2_get_references(
         arxiv_id = arxiv_id.split("v")[0]
     paper_id = f"arxiv:{arxiv_id}"
 
-    url = REFERENCES_URL_TEMPLATE.format(
-        paper_id=paper_id, fields=FIELDS, offset=offset, limit=limit
-    )
-    response = get_with_retries(url)
+    url = REFERENCES_URL_TEMPLATE.format(base_url=BASE_URL, paper_id=paper_id)
+    payload = {"fields": FIELDS, "offset": offset, "limit": limit}
+    response = get_with_retries(url, params=payload)
     result = response.json()
     entries = result["data"]
     total_count = len(result["data"]) + result["offset"]
@@ -143,8 +147,9 @@ def s2_corpus_id_from_arxiv_id(arxiv_id: str) -> int:
     assert isinstance(arxiv_id, str), "Error: Your arxiv_id must be a string"
     if "v" in arxiv_id:
         arxiv_id = arxiv_id.split("v")[0]
-    paper_url = PAPER_URL_TEMPLATE.format(paper_id=f"arxiv:{arxiv_id}", fields="externalIds")
-    response = get_with_retries(paper_url)
+    paper_url = PAPER_URL_TEMPLATE.format(base_url=BASE_URL, paper_id=f"arxiv:{arxiv_id}")
+    payload = {"fields": "externalIds"}
+    response = get_with_retries(paper_url, params=payload)
     result = response.json()
     return int(result["externalIds"]["CorpusId"])
 
@@ -159,8 +164,10 @@ def s2_get_info(arxiv_id: str) -> S2PaperInfo:
     assert isinstance(arxiv_id, str), "Error: Your arxiv_id must be a string"
     if "v" in arxiv_id:
         arxiv_id = arxiv_id.split("v")[0]
-    paper_url = PAPER_URL_TEMPLATE.format(paper_id=f"arxiv:{arxiv_id}", fields=FIELDS)
-    response = get_with_retries(paper_url)
+    paper_id = f"arxiv:{arxiv_id}"
+    payload = {"fields": FIELDS}
+    paper_url = PAPER_URL_TEMPLATE.format(base_url=BASE_URL, paper_id=paper_id)
+    response = get_with_retries(paper_url, params=payload)
     json_data = response.json()
     return S2PaperInfo(
         arxiv_id=json_data.get("externalIds", {}).get("ArXiv"),
@@ -171,3 +178,45 @@ def s2_get_info(arxiv_id: str) -> S2PaperInfo:
         citation_count=int(json_data.get("citationCount", 0)),
         publication_date=str(json_data.get("publicationDate", "")),
     )
+
+
+def s2_search(
+    query: str,
+    offset: int = 0,
+    limit: int = 5,
+    min_citation_count: int = 0,
+    publication_date: Optional[str] = None,
+) -> S2SearchResponse:
+    """
+    Search the S2 corpus for a given query.
+
+    Args:
+        query: The query to search for.
+        offset: The offset to scroll through results. 10 items will be skipped if offset=10. 0 by default.
+        limit: The maximum number of items to return. limit=50 by default.
+        min_citation_count: The minimum citation count to return. 0 by default.
+        publication_date: Restricts results to the given range of publication dates or years (inclusive).
+            Accepts the format <startDate>:<endDate> with each date in YYYY-MM-DD format. None by default.
+    """
+    url = SEARCH_URL_TEMPLATE.format(base_url=BASE_URL)
+    payload = {
+        "query": query,
+        "offset": offset,
+        "limit": limit,
+        "minCitationCount": min_citation_count,
+        "fields": FIELDS,
+    }
+    if publication_date:
+        payload["publicationDateOrYear"] = publication_date
+    response = get_with_retries(url, params=payload, backoff_factor=10.0, num_retries=5)
+    result = response.json()
+    if "data" not in result:
+        return S2SearchResponse(
+            total_count=0,
+            returned_count=0,
+            offset=offset if offset else 0,
+            results=[],
+        )
+    entries = result["data"]
+    total_count = result["total"]
+    return _format_entries(entries, offset if offset else 0, total_count)

{academia_mcp-1.11.4 → academia_mcp-1.11.5}/academia_mcp/utils.py

@@ -42,7 +42,7 @@ def post_with_retries(
 def get_with_retries(
     url: str,
     api_key: Optional[str] = None,
-    timeout: int = 30,
+    timeout: int = 60,
     num_retries: int = 3,
     backoff_factor: float = 3.0,
     params: Optional[Dict[str, Any]] = None,

{academia_mcp-1.11.4 → academia_mcp-1.11.5}/academia_mcp.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: academia-mcp
-Version: 1.11.4
+Version: 1.11.5
 Summary: MCP server that provides different tools to search for scientific publications
 Author-email: Ilya Gusev <phoenixilya@gmail.com>
 Project-URL: Homepage, https://github.com/IlyaGusev/academia_mcp

{academia_mcp-1.11.4 → academia_mcp-1.11.5}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "academia-mcp"
-version = "1.11.4"
+version = "1.11.5"
 description = "MCP server that provides different tools to search for scientific publications"
 readme = "README.md"
 authors = [

{academia_mcp-1.11.4 → academia_mcp-1.11.5}/tests/test_bitflip.py

@@ -35,18 +35,3 @@ async def test_bitflip_score_research_proposals_base() -> None:
     assert scores.proposals[1].strengths is not None
     assert scores.proposals[0].weaknesses is not None
     assert scores.proposals[1].weaknesses is not None
-
-
-async def test_bitflip_score_research_proposals_str() -> None:
-    arxiv_id = "2503.07826"
-    bit = (await extract_bitflip_info(arxiv_id)).bit
-    proposals = await generate_research_proposals(bit=bit, num_proposals=2)
-    scores = await score_research_proposals(proposals)
-    assert scores.proposals
-    assert len(scores.proposals) == 2
-    assert scores.proposals[0].spark is not None
-    assert scores.proposals[1].spark is not None
-    assert scores.proposals[0].strengths is not None
-    assert scores.proposals[1].strengths is not None
-    assert scores.proposals[0].weaknesses is not None
-    assert scores.proposals[1].weaknesses is not None

{academia_mcp-1.11.4 → academia_mcp-1.11.5}/tests/test_s2.py

@@ -3,6 +3,7 @@ from academia_mcp.tools import (
     s2_get_references,
     s2_corpus_id_from_arxiv_id,
     s2_get_info,
+    s2_search,
 )
 
 
@@ -40,3 +41,31 @@ def test_s2_get_info() -> None:
     assert info.citation_count is not None
     assert info.publication_date is not None
     assert info.external_ids["CorpusId"] == 272593138
+
+
+def test_s2_search_base() -> None:
+    result = s2_search("transformers")
+    assert result.total_count >= 1
+    assert "transformers" in str(result.results).lower()
+    assert result.offset == 0
+    assert result.returned_count == 5
+
+
+def test_s2_search_offset() -> None:
+    result = s2_search("transformers", offset=10)
+    assert result.total_count >= 1
+    assert "transformers" in str(result.results).lower()
+    assert result.offset == 10
+    assert result.returned_count == 5
+
+
+def test_s2_search_min_citation_count() -> None:
+    result = s2_search("transformers", min_citation_count=100000)
+    assert result.total_count >= 2 and result.total_count <= 10
+
+
+def test_s2_search_publication_date() -> None:
+    result = s2_search(
+        "transformers", min_citation_count=100000, publication_date="2017-01-01:2017-12-31"
+    )
+    assert result.total_count == 1