mirror of
https://github.com/tcsenpai/agenticSeek.git
synced 2025-06-05 02:25:27 +00:00
Feat : web navigation improvement, better prompting
This commit is contained in:
martin legrand
parent
292623ab52
commit
bdbb590dc4
@ -44,7 +44,7 @@ User: "I need to find the file config.txt and read its contents."
|
||||
|
||||
Assistant: I’ll use file_finder to locate the file:
|
||||
|
||||
```file_finder
|
||||
```file_finder:read
|
||||
config.txt
|
||||
```
|
||||
|
||||
|
||||
@ -15,9 +15,8 @@ class BrowserAgent(Agent):
|
||||
self.tools = {
|
||||
"web_search": searxSearch(),
|
||||
}
|
||||
self.role = "deep research and web search"
|
||||
self.role = "web search, internet, google"
|
||||
self.browser = Browser()
|
||||
self.browser.go_to("https://github.com/")
|
||||
self.search_history = []
|
||||
self.navigable_links = []
|
||||
self.notes = []
|
||||
@ -50,7 +49,7 @@ class BrowserAgent(Agent):
|
||||
{search_choice}
|
||||
Your goal is to find accurate and complete information to satisfy the user’s request.
|
||||
User request: {user_prompt}
|
||||
To proceed, choose a relevant link from the search results. Announce your choice by saying: "I want to navigate to <link>."
|
||||
To proceed, choose a relevant link from the search results. Announce your choice by saying: "I want to navigate to <link>"
|
||||
Do not explain your choice.
|
||||
"""
|
||||
|
||||
@ -58,27 +57,44 @@ class BrowserAgent(Agent):
|
||||
remaining_links = self.get_unvisited_links()
|
||||
remaining_links_text = remaining_links if remaining_links is not None else "No links remaining, proceed with a new search."
|
||||
return f"""
|
||||
\nYou are currently browsing the web. Not the user, you are the browser.
|
||||
|
||||
Page content:
|
||||
You are a web browser.
|
||||
You are currently on this webpage:
|
||||
{page_text}
|
||||
|
||||
You can navigate to these links:
|
||||
You can navigate to these navigation links:
|
||||
{remaining_links}
|
||||
|
||||
You must choose a link (write it down) to navigate to, or go back.
|
||||
For exemple you can say: i want to go to www.wikipedia.org/cats
|
||||
Your task:
|
||||
1. Decide if the current page answers the user’s query: {user_prompt}
|
||||
- If it does, take notes of the useful information, write down source, link or reference, then move to a new page.
|
||||
- If it does and you are 100% certain that it provide a definive answer, say REQUEST_EXIT
|
||||
- If it doesn’t, say: Error: This page does not answer the user’s query then go back or navigate to another link.
|
||||
2. Navigate by either:
|
||||
- Navigate to a navigation links (write the full URL, e.g., www.example.com/cats).
|
||||
- If no link seems helpful, say: GO_BACK.
|
||||
|
||||
Recap of note taking:
|
||||
If useful -> Note: [Briefly summarize the key information that answers the user’s query.]
|
||||
Do not write "The page talk about ...", write your finding on the page and how they contribute to an answer.
|
||||
If not useful -> Error: [Explain why the page doesn’t help.]
|
||||
|
||||
Example 1 (useful page, no need of going futher):
|
||||
Note: According to karpathy site (https://karpathy.github.io/) LeCun net is the earliest real-world application of a neural net"
|
||||
No link seem useful to provide futher information. GO_BACK
|
||||
|
||||
Follow up with a summary of the page content (of the current page, not of the link), for example:
|
||||
Summary: According to https://karpathy.github.io/ LeCun net is the earliest real-world application of a neural net"
|
||||
The summary should include any useful finding that are useful in answering user query.
|
||||
If a website does not have usefull information say Error, for exemple:
|
||||
Error: This forum does not discus anything that can answer the user query
|
||||
Be short, concise, direct.
|
||||
Example 2 (not useful, but related link):
|
||||
Error: This forum reddit.com/welcome does not discuss anything related to the user’s query.
|
||||
There is a link that could lead to the information, I want to navigate to http://reddit.com/r/locallama
|
||||
|
||||
If no link seem appropriate, please say "GO_BACK".
|
||||
Remember, you seek the information the user want.
|
||||
The user query was : {user_prompt}
|
||||
Example 3 (not useful, no related links):
|
||||
Error: x.com does not discuss anything related to the user’s query and no navigation link are usefull
|
||||
GO_BACK
|
||||
|
||||
Example 3 (not useful, no related links):
|
||||
Note: I found on github.com that the creator of agenticSeek is fosowl.
|
||||
Given this information, given this I should exit the web browser. REQUEST_EXIT
|
||||
|
||||
Remember, the user asked: {user_prompt}
|
||||
"""
|
||||
|
||||
def llm_decide(self, prompt):
|
||||
@ -122,11 +138,11 @@ class BrowserAgent(Agent):
|
||||
def save_notes(self, text):
|
||||
lines = text.split('\n')
|
||||
for line in lines:
|
||||
if "summary" in line.lower():
|
||||
if "note" in line.lower():
|
||||
self.notes.append(line)
|
||||
|
||||
def conclude_prompt(self, user_query):
|
||||
annotated_notes = [f"{i+1}: {note.lower().replace('summary:', '')}" for i, note in enumerate(self.notes)]
|
||||
annotated_notes = [f"{i+1}: {note.lower().replace('note:', '')}" for i, note in enumerate(self.notes)]
|
||||
search_note = '\n'.join(annotated_notes)
|
||||
print("AI research notes:\n", search_note)
|
||||
return f"""
|
||||
@ -174,7 +190,6 @@ class BrowserAgent(Agent):
|
||||
self.memory.push('user', prompt)
|
||||
answer, reasoning = self.llm_request(prompt)
|
||||
pretty_print(answer, color="output")
|
||||
speech_module.speak(answer)
|
||||
return answer, reasoning
|
||||
|
||||
if __name__ == "__main__":
|
||||
|
||||
@ -14,7 +14,7 @@ class FileAgent(Agent):
|
||||
"file_finder": FileFinder(),
|
||||
"bash": BashInterpreter()
|
||||
}
|
||||
self.role = "files operations"
|
||||
self.role = "find, read, write, edit files"
|
||||
|
||||
def process(self, prompt, speech_module) -> str:
|
||||
complete = False
|
||||
|
||||
@ -13,6 +13,7 @@ import markdownify
|
||||
import logging
|
||||
import sys
|
||||
import re
|
||||
from urllib.parse import urlparse
|
||||
|
||||
class Browser:
|
||||
def __init__(self, headless=False, anticaptcha_install=False):
|
||||
@ -58,7 +59,8 @@ class Browser:
|
||||
os.path.join(os.environ.get("LOCALAPPDATA", ""), "Google\\Chrome\\Application\\chrome.exe") # User install
|
||||
]
|
||||
elif sys.platform.startswith("darwin"): # macOS
|
||||
paths = ["/Applications/Google Chrome.app/Contents/MacOS/Google Chrome", "/Applications/Google Chrome Beta.app/Contents/MacOS/Google Chrome Beta"]
|
||||
paths = ["/Applications/Google Chrome.app/Contents/MacOS/Google Chrome",
|
||||
"/Applications/Google Chrome Beta.app/Contents/MacOS/Google Chrome Beta"]
|
||||
else: # Linux
|
||||
paths = ["/usr/bin/google-chrome", "/usr/bin/chromium-browser", "/usr/bin/chromium"]
|
||||
|
||||
@ -79,19 +81,6 @@ class Browser:
|
||||
self.logger.error(f"Error navigating to {url}: {str(e)}")
|
||||
return False
|
||||
|
||||
def is_sentence(self, text):
|
||||
"""Check if the text qualifies as a meaningful sentence or contains important error codes."""
|
||||
text = text.strip()
|
||||
error_codes = ["404", "403", "500", "502", "503"]
|
||||
if any(code in text for code in error_codes):
|
||||
return True
|
||||
words = text.split()
|
||||
word_count = len(words)
|
||||
has_punctuation = text.endswith(('.', '!', '?'))
|
||||
is_long_enough = word_count > 5
|
||||
has_letters = any(word.isalpha() for word in words)
|
||||
return (word_count >= 5 and (has_punctuation or is_long_enough) and has_letters)
|
||||
|
||||
def is_sentence(self, text):
|
||||
"""Check if the text qualifies as a meaningful sentence or contains important error codes."""
|
||||
text = text.strip()
|
||||
@ -118,10 +107,8 @@ class Browser:
|
||||
lines = (line.strip() for line in text.splitlines())
|
||||
chunks = (phrase.strip() for line in lines for phrase in line.split(" "))
|
||||
text = "\n".join(chunk for chunk in chunks if chunk and self.is_sentence(chunk))
|
||||
|
||||
markdown_text = markdownify.markdownify(text, heading_style="ATX")
|
||||
|
||||
return markdown_text
|
||||
#markdown_text = markdownify.markdownify(text, heading_style="ATX")
|
||||
return "[Start of page]\n" + text + "\n[End of page]"
|
||||
except Exception as e:
|
||||
self.logger.error(f"Error getting text: {str(e)}")
|
||||
return None
|
||||
@ -142,6 +129,22 @@ class Browser:
|
||||
if essential_params:
|
||||
return f"{base_url}?{'&'.join(essential_params)}"
|
||||
return base_url
|
||||
|
||||
def is_link_valid(self, url):
|
||||
"""Check if a URL is a valid link (page, not related to icon or metadata)."""
|
||||
if len(url) > 64:
|
||||
return False
|
||||
parsed_url = urlparse(url)
|
||||
if not parsed_url.scheme or not parsed_url.netloc:
|
||||
return False
|
||||
if re.search(r'/\d+$', parsed_url.path):
|
||||
return False
|
||||
image_extensions = ['.jpg', '.jpeg', '.png', '.gif', '.bmp', '.tiff', '.webp']
|
||||
metadata_extensions = ['.ico', '.xml', '.json', '.rss', '.atom']
|
||||
for ext in image_extensions + metadata_extensions:
|
||||
if url.lower().endswith(ext):
|
||||
return False
|
||||
return True
|
||||
|
||||
def get_navigable(self):
|
||||
"""Get all navigable links on the current page."""
|
||||
@ -159,7 +162,7 @@ class Browser:
|
||||
})
|
||||
|
||||
self.logger.info(f"Found {len(links)} navigable links")
|
||||
return [self.clean_url(link['url']) for link in links if (link['is_displayed'] == True and len(link['url']) < 64)]
|
||||
return [self.clean_url(link['url']) for link in links if (link['is_displayed'] == True and self.is_link_valid(link['url']))]
|
||||
except Exception as e:
|
||||
self.logger.error(f"Error getting navigable links: {str(e)}")
|
||||
return []
|
||||
@ -225,7 +228,7 @@ if __name__ == "__main__":
|
||||
browser = Browser(headless=False)
|
||||
|
||||
try:
|
||||
browser.go_to("https://www.seoul.co.kr/")
|
||||
browser.go_to("https://github.com/Fosowl/agenticSeek")
|
||||
text = browser.get_text()
|
||||
print("Page Text in Markdown:")
|
||||
print(text)
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user