diff --git a/src/deep_research.py b/src/deep_research.py index 204de7217..375d8d8ab 100644 --- a/src/deep_research.py +++ b/src/deep_research.py @@ -16,7 +16,8 @@ from typing import Callable, Dict, List, Optional, Set from src.research_utils import strip_thinking, is_low_quality -from src.goal_based_extractor import EXTRACTOR_PROMPT +from src.goal_based_extractor import EXTRACTOR_SYSTEM +from src.prompt_security import untrusted_context_message logger = logging.getLogger(__name__) @@ -625,11 +626,12 @@ class DeepResearcher: else: content = truncated - prompt = EXTRACTOR_PROMPT.format(webpage_content=content, goal=question) - try: response = await self._llm( - [{"role": "user", "content": prompt}], + [ + {"role": "user", "content": EXTRACTOR_SYSTEM.format(goal=question)}, + untrusted_context_message("webpage", content), + ], temperature=0.2, max_tokens=2048, timeout=self.extraction_timeout, diff --git a/src/goal_based_extractor.py b/src/goal_based_extractor.py index 219214466..cd43b96ac 100644 --- a/src/goal_based_extractor.py +++ b/src/goal_based_extractor.py @@ -3,22 +3,18 @@ Goal-based content extraction prompt inspired by Alibaba Tongyi DeepResearch. """ -EXTRACTOR_PROMPT = """Please process the following webpage content and user goal to extract relevant information: +EXTRACTOR_SYSTEM = """Extract relevant information from a webpage for a given research goal. -## **Webpage Content** -{webpage_content} +Goal: {goal} -## **User Goal** -{goal} +Task guidelines: +1. Locate the specific sections directly related to the goal within the provided webpage content. +2. Identify and extract the most relevant information; output full original context where possible, up to three or more paragraphs. +3. Organize into a concise paragraph with logical flow, judging each piece of information's contribution to the goal. -## **Task Guidelines** -1. **Content Scanning for Rational**: Locate the **specific sections/data** directly related to the user's goal within the webpage content -2. **Key Extraction for Evidence**: Identify and extract the **most relevant information** from the content, you never miss any important information, output the **full original context** of the content as far as possible, it can be more than three paragraphs. -3. **Summary Output for Summary**: Organize into a concise paragraph with logical flow, prioritizing clarity and judge the contribution of the information to the goal. +Respond in JSON with exactly these fields: "rational", "evidence", "summary". -**Final Output Format using JSON format has "rational", "evidence", "summary" fields** - -Example output: +Example: {{ "rational": "This section discusses X which directly relates to the goal of understanding Y", "evidence": "Full quotes and context from the page...",