From 7d47ca54be6e6b04a45d54d764f612421f82e2d1 Mon Sep 17 00:00:00 2001
From: Roberto <roberto.musso@hpecds.com>
Date: Tue, 12 May 2026 16:40:20 +0200
Subject: [PATCH] feat(api): emit Langfuse generation traces for folder indexer

---
 app/core/folder_indexer.py | 39 +++++++++++++++++++++++++++++++++-----
 1 file changed, 34 insertions(+), 5 deletions(-)

diff --git a/app/core/folder_indexer.py b/app/core/folder_indexer.py
index 4a070db..43af1c5 100644
--- a/app/core/folder_indexer.py
+++ b/app/core/folder_indexer.py
@@ -12,6 +12,7 @@ from docx import Document as DocxDocument
 from app.core.langfuse_client import (
     compile_prompt,
     extract_usage,
+    get_langfuse,
     get_prompt_or_fallback,
 )
 from app.core.llm import get_llm
@@ -55,7 +56,7 @@ async def _llm_vision(messages: list) -> object:
     return await llm.ainvoke(messages)
 
 
-async def summarize_image(*, image_b64: str, mime: str) -> IndexResult:
+async def summarize_image(*, image_b64: str, mime: str, file_name: str | None = None) -> IndexResult:
     """Return a compact summary of an image file using vision.
 
     Parameters
@@ -64,6 +65,8 @@ async def summarize_image(*, image_b64: str, mime: str) -> IndexResult:
         Base64-encoded image bytes.
     mime:
         MIME type of the image, e.g. ``"image/png"``.
+    file_name:
+        Optional file name, attached to the Langfuse trace as input metadata.
     """
     template, prompt_obj = get_prompt_or_fallback("folder_file_summary_image", _IMAGE_FALLBACK)
     messages = [
@@ -73,8 +76,21 @@ async def summarize_image(*, image_b64: str, mime: str) -> IndexResult:
             {"type": "image_url", "image_url": {"url": f"data:{mime};base64,{image_b64}"}},
         ]),
     ]
-    response = await _llm_vision(messages)
-    usage = extract_usage(response)
+    lf = get_langfuse()
+    if lf is not None:
+        with lf.start_as_current_observation(
+            as_type="generation",
+            name="folder-summarize-image",
+            model="gpt-4o-mini",
+            prompt=prompt_obj,
+            input={"file_name": file_name, "mime": mime},
+        ) as gen:
+            response = await _llm_vision(messages)
+            usage = extract_usage(response)
+            gen.update(output=response.content, usage_details=usage)
+    else:
+        response = await _llm_vision(messages)
+        usage = extract_usage(response)
     summary = (response.content or "").strip()[:500]
     return IndexResult(summary=summary, tokens_used=usage.get("total", 0))
 
@@ -98,8 +114,21 @@ async def summarize_text(*, content: str, ext: str, name: str) -> IndexResult:
         SystemMessage(content=compiled),
         HumanMessage(content="Summarise this file."),
     ]
-    response = await _llm_text(messages)
-    usage = extract_usage(response)
+    lf = get_langfuse()
+    if lf is not None:
+        with lf.start_as_current_observation(
+            as_type="generation",
+            name="folder-summarize-text",
+            model="gpt-4o-mini",
+            prompt=prompt_obj,
+            input={"file_name": name, "ext": ext, "content_chars": len(truncated)},
+        ) as gen:
+            response = await _llm_text(messages)
+            usage = extract_usage(response)
+            gen.update(output=response.content, usage_details=usage)
+    else:
+        response = await _llm_text(messages)
+        usage = extract_usage(response)
     summary = (response.content or "").strip()[:500]
     return IndexResult(summary=summary, tokens_used=usage.get("total", 0))