Add AI summarization

README.md

@@ -92,6 +92,12 @@ BROWSER_TIMEOUT=30000
 MAX_CONTENT_LENGTH=10000
 GENERATE_SUMMARY=true
 
+# OpenAI/OpenRouter
+OPENAI_API_KEY=your_key_here
+OPENAI_URL=https://openrouter.ai/api/v1/chat/completions
+OPENAI_MODEL=gpt-4o-mini
+OPENAI_TIMEOUT=30
+
 # YouTube
 YOUTUBE_LANGUAGE=en
 

config.py

@@ -27,6 +27,12 @@ class Config:
     MAX_CONTENT_LENGTH = int(os.getenv("MAX_CONTENT_LENGTH", "10000"))  # Max chars
     GENERATE_SUMMARY = os.getenv("GENERATE_SUMMARY", "true").lower() == "true"
 
+    # OpenAI/OpenRouter settings
+    OPENAI_API_KEY = os.getenv("OPENAI_API_KEY", "")
+    OPENAI_URL = os.getenv("OPENAI_URL", "https://api.openai.com/v1/chat/completions")
+    OPENAI_MODEL = os.getenv("OPENAI_MODEL", "gpt-4o-mini")
+    OPENAI_TIMEOUT = int(os.getenv("OPENAI_TIMEOUT", "30"))
+
     # YouTube settings
     YOUTUBE_LANGUAGE = os.getenv("YOUTUBE_LANGUAGE", "en")
 

main.py

@@ -23,6 +23,7 @@ from extractors.blog_extractor import BlogExtractor
 from extractors.instagram_extractor import InstagramExtractor
 from obsidian_writer import ObsidianWriter
 from config import Config
+from summarizer import summarize_text, SummarizationError
 
 
 def detect_source_type(url: str) -> str:
@@ -102,6 +103,22 @@ def main():
         print("❌ No content could be extracted")
         sys.exit(1)
 
+    # Generate AI summary + key points
+    if args.summarize or Config.GENERATE_SUMMARY:
+        source_text = "\n\n".join(
+            part for part in [content.get("description", ""), content.get("content", "")]
+            if part
+        ).strip()
+        if source_text:
+            try:
+                summary_result = summarize_text(source_text, max_points=3)
+                if summary_result.get("summary"):
+                    content["description"] = summary_result["summary"]
+                if summary_result.get("key_points"):
+                    content["key_points"] = summary_result["key_points"]
+            except SummarizationError as e:
+                print(f"⚠️  Summarization failed: {e}")
+
     # Generate output filename
     if args.output:
         filename = args.output

summarizer.py

@@ -0,0 +1,84 @@
+"""
+OpenAI/OpenRouter summarizer utility.
+
+Uses OPENAI_API_KEY and OPENAI_URL from environment (via Config).
+"""
+
+from __future__ import annotations
+
+import json
+from typing import Dict, List
+
+import requests
+
+from config import Config
+
+
+class SummarizationError(RuntimeError):
+    """Raised when summarization fails."""
+
+
+def summarize_text(text: str, max_points: int = 3) -> Dict[str, List[str] | str]:
+    """
+    Summarize text into a short summary and key points.
+
+    Returns:
+        {
+            "summary": "string",
+            "key_points": ["point 1", "point 2", ...]
+        }
+    """
+    if not text or not text.strip():
+        return {"summary": "", "key_points": []}
+
+    if not Config.OPENAI_API_KEY:
+        raise SummarizationError("OPENAI_API_KEY is not set")
+
+    payload = {
+        "model": Config.OPENAI_MODEL,
+        "messages": [
+            {
+                "role": "system",
+                "content": (
+                    "You are a precise summarizer. Return JSON only with keys "
+                    "`summary` and `key_points` (array of strings). Do not add extra keys."
+                ),
+            },
+            {
+                "role": "user",
+                "content": (
+                    "Summarize the following content in 2-4 sentences and provide "
+                    f"{max_points} key points.\n\n"
+                    f"CONTENT:\n{text}"
+                ),
+            },
+        ],
+        "temperature": 0.2,
+        "max_tokens": 400,
+    }
+
+    headers = {
+        "Authorization": f"Bearer {Config.OPENAI_API_KEY}",
+        "Content-Type": "application/json",
+    }
+
+    try:
+        response = requests.post(
+            Config.OPENAI_URL,
+            headers=headers,
+            json=payload,
+            timeout=Config.OPENAI_TIMEOUT,
+        )
+        response.raise_for_status()
+        data = response.json()
+    except Exception as exc:
+        raise SummarizationError(f"Request failed: {exc}") from exc
+
+    try:
+        content = data["choices"][0]["message"]["content"].strip()
+        result = json.loads(content)
+        summary = result.get("summary", "").strip()
+        key_points = [p.strip() for p in result.get("key_points", []) if p.strip()]
+        return {"summary": summary, "key_points": key_points}
+    except Exception as exc:
+        raise SummarizationError(f"Invalid response format: {exc}") from exc