diff --git a/plugins/omi-crossref-app/Procfile b/plugins/omi-crossref-app/Procfile
new file mode 100644
index 00000000000..8f0921228d8
--- /dev/null
+++ b/plugins/omi-crossref-app/Procfile
@@ -0,0 +1 @@
+web: uvicorn main:app --host 0.0.0.0 --port ${PORT:-8080}
diff --git a/plugins/omi-crossref-app/README.md b/plugins/omi-crossref-app/README.md
new file mode 100644
index 00000000000..6fd4fe279e8
--- /dev/null
+++ b/plugins/omi-crossref-app/README.md
@@ -0,0 +1,17 @@
+# Omi Crossref App
+
+A no-auth Crossref integration app for Omi chat tools.
+
+## Tools
+- `search_crossref_works`: search scholarly works by query
+- `get_crossref_work`: fetch work details by DOI
+- `get_crossref_works_by_author`: list recent works by author name
+
+## Run locally
+```bash
+pip install -r requirements.txt
+uvicorn main:app --reload --host 0.0.0.0 --port 8080
+```
+
+## Deploy
+Deploy this folder on Railway/Heroku-style hosts using the included `Procfile`/`railway.toml`.
diff --git a/plugins/omi-crossref-app/main.py b/plugins/omi-crossref-app/main.py
new file mode 100644
index 00000000000..e5d86d2f631
--- /dev/null
+++ b/plugins/omi-crossref-app/main.py
@@ -0,0 +1,243 @@
+import html
+from typing import Any
+from urllib.parse import quote
+
+import httpx
+from fastapi import FastAPI
+
+from models import AuthorWorksInput, ChatToolResponse, GetWorkInput, SearchWorksInput
+
+CROSSREF_BASE = "https://api.crossref.org"
+TIMEOUT = 20.0
+
+app = FastAPI(
+    title="Crossref Omi Integration",
+    description="No-auth Crossref chat tools for paper metadata search and lookup",
+    version="1.0.0",
+)
+
+
+def clamp_max_results(value: int) -> int:
+    return max(1, min(10, value))
+
+
+def clean(text: Any) -> str:
+    if text is None:
+        return ""
+    return html.unescape(str(text)).strip()
+
+
+def extract_year(item: dict[str, Any]) -> str:
+    for key in ("published-print", "published-online", "issued"):
+        date_parts = (item.get(key) or {}).get("date-parts", [])
+        if date_parts and date_parts[0]:
+            return clean(date_parts[0][0])
+    return ""
+
+
+async def crossref_get(path: str, params: dict[str, Any]) -> dict[str, Any]:
+    async with httpx.AsyncClient(timeout=TIMEOUT) as client:
+        response = await client.get(f"{CROSSREF_BASE}{path}", params=params)
+        response.raise_for_status()
+        return response.json()
+
+
+@app.get("/health")
+async def health():
+    return {"status": "ok"}
+
+
+@app.get("/tools")
+async def tools():
+    return {
+        "tools": [
+            {
+                "name": "search_crossref_works",
+                "description": "Search scholarly works by keyword via Crossref",
+                "parameters": {
+                    "query": {"type": "string", "description": "Search keyword(s)"},
+                    "max_results": {
+                        "type": "integer",
+                        "description": "Number of results (1-10)",
+                        "default": 5,
+                    },
+                },
+            },
+            {
+                "name": "get_crossref_work",
+                "description": "Get details for a specific work by DOI",
+                "parameters": {
+                    "doi": {
+                        "type": "string",
+                        "description": "DOI, e.g. 10.1038/nphys1170",
+                    }
+                },
+            },
+            {
+                "name": "get_crossref_works_by_author",
+                "description": "Find recent works for an author name",
+                "parameters": {
+                    "author": {"type": "string", "description": "Author name"},
+                    "max_results": {
+                        "type": "integer",
+                        "description": "Number of results (1-10)",
+                        "default": 5,
+                    },
+                },
+            },
+        ]
+    }
+
+
+@app.get("/.well-known/omi-tools.json")
+async def get_omi_tools_manifest():
+    return {
+        "tools": [
+            {
+                "name": "search_crossref_works",
+                "description": "Search scholarly works by keyword via Crossref",
+                "endpoint": "/tools/search_crossref_works",
+                "method": "POST",
+                "parameters": {
+                    "properties": {
+                        "query": {"type": "string", "description": "Search keyword(s)"},
+                        "max_results": {
+                            "type": "integer",
+                            "description": "Number of results (1-10)",
+                            "default": 5,
+                        },
+                    },
+                    "required": ["query"],
+                },
+                "auth_required": False,
+                "status_message": "Searching Crossref...",
+            },
+            {
+                "name": "get_crossref_work",
+                "description": "Get details for a specific work by DOI",
+                "endpoint": "/tools/get_crossref_work",
+                "method": "POST",
+                "parameters": {
+                    "properties": {
+                        "doi": {
+                            "type": "string",
+                            "description": "DOI, e.g. 10.1038/nphys1170",
+                        }
+                    },
+                    "required": ["doi"],
+                },
+                "auth_required": False,
+                "status_message": "Fetching Crossref work...",
+            },
+            {
+                "name": "get_crossref_works_by_author",
+                "description": "Find recent works for an author name",
+                "endpoint": "/tools/get_crossref_works_by_author",
+                "method": "POST",
+                "parameters": {
+                    "properties": {
+                        "author": {"type": "string", "description": "Author name"},
+                        "max_results": {
+                            "type": "integer",
+                            "description": "Number of results (1-10)",
+                            "default": 5,
+                        },
+                    },
+                    "required": ["author"],
+                },
+                "auth_required": False,
+                "status_message": "Fetching author works...",
+            },
+        ]
+    }
+
+
+@app.post("/tools/search_crossref_works", response_model=ChatToolResponse)
+async def search_crossref_works(payload: SearchWorksInput):
+    query = payload.query.strip()
+    if len(query) < 2:
+        return ChatToolResponse(error="Query must be at least 2 characters.")
+    limited = clamp_max_results(payload.max_results)
+    try:
+        payload = await crossref_get(
+            "/works",
+            {"query": query, "rows": limited, "sort": "relevance", "order": "desc"},
+        )
+    except Exception as exc:
+        return ChatToolResponse(error=f"Crossref request failed: {exc}")
+    items = payload.get("message", {}).get("items", [])
+    if not items:
+        return ChatToolResponse(result=f"No Crossref results found for '{query}'.")
+
+    lines = [f"Top {len(items)} Crossref results for '{query}':"]
+    for idx, item in enumerate(items, 1):
+        title = clean((item.get("title") or ["Untitled"])[0])
+        doi = clean(item.get("DOI"))
+        year = extract_year(item)
+        lines.append(f"{idx}. {title} ({year})")
+        lines.append(f"   DOI: {doi}")
+    return ChatToolResponse(result="\n".join(lines))
+
+
+@app.post("/tools/get_crossref_work", response_model=ChatToolResponse)
+async def get_crossref_work(payload: GetWorkInput):
+    normalized = payload.doi.strip()
+    if "/" not in normalized:
+        return ChatToolResponse(error="Invalid DOI format. Example: 10.1038/nphys1170")
+    if ".." in normalized:
+        return ChatToolResponse(error="Invalid DOI value.")
+
+    try:
+        payload = await crossref_get(f"/works/{quote(normalized, safe='')}", {})
+    except Exception as exc:
+        return ChatToolResponse(error=f"Crossref request failed: {exc}")
+    item = payload.get("message", {})
+    title = clean((item.get("title") or ["Untitled"])[0])
+    publisher = clean(item.get("publisher"))
+    doi_out = clean(item.get("DOI"))
+    url = clean(item.get("URL"))
+    abstract = clean(item.get("abstract"))
+    year = extract_year(item)
+
+    parts = [
+        f"Title: {title}",
+        f"DOI: {doi_out}",
+        f"Year: {year}",
+        f"Publisher: {publisher}",
+        f"URL: {url}",
+    ]
+    if abstract:
+        parts.append(f"Abstract: {abstract[:1200]}")
+    return ChatToolResponse(result="\n".join(parts))
+
+
+@app.post("/tools/get_crossref_works_by_author", response_model=ChatToolResponse)
+async def get_crossref_works_by_author(payload: AuthorWorksInput):
+    author = payload.author.strip()
+    if len(author) < 2:
+        return ChatToolResponse(error="Author must be at least 2 characters.")
+    limited = clamp_max_results(payload.max_results)
+    try:
+        payload = await crossref_get(
+            "/works",
+            {
+                "query.author": author,
+                "rows": limited,
+                "sort": "published",
+                "order": "desc",
+            },
+        )
+    except Exception as exc:
+        return ChatToolResponse(error=f"Crossref request failed: {exc}")
+    items = payload.get("message", {}).get("items", [])
+    if not items:
+        return ChatToolResponse(result=f"No recent works found for author '{author}'.")
+
+    lines = [f"Recent works for '{author}':"]
+    for idx, item in enumerate(items, 1):
+        title = clean((item.get("title") or ["Untitled"])[0])
+        doi = clean(item.get("DOI"))
+        year = extract_year(item)
+        lines.append(f"{idx}. {title} ({year})")
+        lines.append(f"   DOI: {doi}")
+    return ChatToolResponse(result="\n".join(lines))
diff --git a/plugins/omi-crossref-app/models.py b/plugins/omi-crossref-app/models.py
new file mode 100644
index 00000000000..c5b5d1062c8
--- /dev/null
+++ b/plugins/omi-crossref-app/models.py
@@ -0,0 +1,22 @@
+from typing import Optional
+
+from pydantic import BaseModel
+
+
+class ChatToolResponse(BaseModel):
+    result: Optional[str] = None
+    error: Optional[str] = None
+
+
+class SearchWorksInput(BaseModel):
+    query: str
+    max_results: int = 5
+
+
+class GetWorkInput(BaseModel):
+    doi: str
+
+
+class AuthorWorksInput(BaseModel):
+    author: str
+    max_results: int = 5
diff --git a/plugins/omi-crossref-app/railway.toml b/plugins/omi-crossref-app/railway.toml
new file mode 100644
index 00000000000..d0c042239af
--- /dev/null
+++ b/plugins/omi-crossref-app/railway.toml
@@ -0,0 +1,9 @@
+[build]
+builder = "NIXPACKS"
+
+[deploy]
+startCommand = "uvicorn main:app --host 0.0.0.0 --port $PORT"
+healthcheckPath = "/health"
+healthcheckTimeout = 100
+restartPolicyType = "ON_FAILURE"
+restartPolicyMaxRetries = 10
diff --git a/plugins/omi-crossref-app/requirements.txt b/plugins/omi-crossref-app/requirements.txt
new file mode 100644
index 00000000000..0a9315fdb44
--- /dev/null
+++ b/plugins/omi-crossref-app/requirements.txt
@@ -0,0 +1,4 @@
+fastapi==0.104.1
+uvicorn==0.24.0
+httpx==0.27.0
+pydantic==2.5.2