diff --git a/src/skillspector/llm_utils.py b/src/skillspector/llm_utils.py
index d1c51040..d698d66d 100644
--- a/src/skillspector/llm_utils.py
+++ b/src/skillspector/llm_utils.py
@@ -30,6 +30,11 @@
 
 from __future__ import annotations
 
+import asyncio
+import concurrent.futures
+from collections.abc import Coroutine
+from typing import Any
+
 from langchain_core.language_models.chat_models import BaseChatModel
 from langchain_core.messages import BaseMessage
 
@@ -106,3 +111,29 @@ def chat_completion(prompt: str, *, model: str | None = None) -> str:
     if not isinstance(response, BaseMessage):
         raise TypeError(f"Expected BaseMessage from chat model, got {type(response).__name__}")
     return str(response.text)
+
+
+def run_async(coroutine: Coroutine) -> Any:
+    """
+    Run an async coroutine in a synchronous context, even if there's already a running event loop.
+
+    This function safely handles nested event loop scenarios (e.g. Jupyter Notebooks, FastAPI,
+    LangGraph Studio) by offloading the coroutine execution to a separate thread with its own
+    event loop when a running loop is detected.
+
+    Args:
+        coroutine: The async coroutine to run
+
+    Returns:
+        The result of the coroutine execution
+
+    Raises:
+        Any exception raised by the coroutine is re-raised as-is
+    """
+    try:
+        asyncio.get_running_loop()
+    except RuntimeError:
+        return asyncio.run(coroutine)
+
+    with concurrent.futures.ThreadPoolExecutor(max_workers=1) as executor:
+        return executor.submit(asyncio.run, coroutine).result()
diff --git a/src/skillspector/nodes/analyzers/semantic_developer_intent.py b/src/skillspector/nodes/analyzers/semantic_developer_intent.py
index a3a54be2..e621141b 100644
--- a/src/skillspector/nodes/analyzers/semantic_developer_intent.py
+++ b/src/skillspector/nodes/analyzers/semantic_developer_intent.py
@@ -22,10 +22,9 @@
 
 from __future__ import annotations
 
-import asyncio
-
 from skillspector.constants import _SKILLSPECTOR_DEFAULT_MODEL, MODEL_CONFIG
 from skillspector.llm_analyzer_base import LLMAnalyzerBase
+from skillspector.llm_utils import run_async
 from skillspector.logging_config import get_logger
 from skillspector.state import AnalyzerNodeResponse, SkillspectorState
 
@@ -176,7 +175,7 @@ def node(state: SkillspectorState) -> AnalyzerNodeResponse:
         prompt = ANALYZER_PROMPT.format(manifest_section=_format_manifest(manifest))
         analyzer = LLMAnalyzerBase(base_prompt=prompt, model=model)
         batches = analyzer.get_batches(sorted(file_cache), file_cache)
-        results = asyncio.run(analyzer.arun_batches(batches))
+        results = run_async(analyzer.arun_batches(batches))
         findings = analyzer.collect_findings(results)
         logger.info("%s: %d findings", ANALYZER_ID, len(findings))
         return {"findings": findings}
diff --git a/src/skillspector/nodes/analyzers/semantic_quality_policy.py b/src/skillspector/nodes/analyzers/semantic_quality_policy.py
index 3140334e..f22a0005 100644
--- a/src/skillspector/nodes/analyzers/semantic_quality_policy.py
+++ b/src/skillspector/nodes/analyzers/semantic_quality_policy.py
@@ -22,10 +22,9 @@
 
 from __future__ import annotations
 
-import asyncio
-
 from skillspector.constants import _SKILLSPECTOR_DEFAULT_MODEL
 from skillspector.llm_analyzer_base import LLMAnalyzerBase
+from skillspector.llm_utils import run_async
 from skillspector.logging_config import get_logger
 from skillspector.state import AnalyzerNodeResponse, SkillspectorState
 
@@ -145,7 +144,7 @@ def node(state: SkillspectorState) -> AnalyzerNodeResponse:
     try:
         analyzer = LLMAnalyzerBase(base_prompt=ANALYZER_PROMPT, model=model)
         batches = analyzer.get_batches(files, file_cache)
-        results = asyncio.run(analyzer.arun_batches(batches))
+        results = run_async(analyzer.arun_batches(batches))
         findings = analyzer.collect_findings(results)
         logger.info("%s: %d findings", ANALYZER_ID, len(findings))
         return {"findings": findings}
diff --git a/src/skillspector/nodes/meta_analyzer.py b/src/skillspector/nodes/meta_analyzer.py
index a1fff859..ac2b0ab3 100644
--- a/src/skillspector/nodes/meta_analyzer.py
+++ b/src/skillspector/nodes/meta_analyzer.py
@@ -22,7 +22,6 @@
 
 from __future__ import annotations
 
-import asyncio
 import json
 from typing import Literal
 
@@ -33,6 +32,7 @@
     LLMAnalyzerBase,
     estimate_tokens,
 )
+from skillspector.llm_utils import run_async
 from skillspector.logging_config import get_logger
 from skillspector.models import Finding
 from skillspector.nodes.analyzers.pattern_defaults import (
@@ -532,7 +532,7 @@ def meta_analyzer(state: SkillspectorState) -> MetaAnalyzerResponse:
             model,
         )
 
-        batch_results = asyncio.run(analyzer.arun_batches(batches, metadata_text=metadata_text))
+        batch_results = run_async(analyzer.arun_batches(batches, metadata_text=metadata_text))
 
         if len(batch_results) < len(batches):
             # Some batches never returned. A finding the LLM never saw has no