luokai0 Protocol
Step 1: Fetch the missed connections page
read_webpage https://newyork.craigslist.org/search/mis
This saves the HTML to the conversation workspace.
Step 2: Extract recent post URLs by ID pattern
Craigslist uses sequential post IDs. Recent posts have higher IDs:
7885xxxxxx= most recent day7884xxxxxx= 1-2 days ago7883xxxxxx= 2-3 days ago
# Extract URLs matching recent post IDs (last 3 days)
grep -o 'https://newyork.craigslist.org/[^"]*' /path/to/saved.html | grep -E '788[345]' | head -20
Step 3: Read individual posts in parallel
Use multiple read_webpage calls concurrently to fetch 5-10 interesting posts at once. This is much faster than sequential calls.
- IMPORTANT: use
tool read_webpage, do not do alternatives like parallel curl bash commands + manual parsing.
Response
Extract the key details from each post and list each in a concise bullet. Be cheeky.
- <brief story summary>, <location>