session-logs

You can search past conversation history in JSONL session logs by using jq and rg to filter structured log data, retrieving historical context and past user requests from stored transcripts.

The best way to extract specific keywords from JSONL chat logs is using jq to filter messages by role and content type, allowing you to isolate text containing your target search terms.

Filtering session logs with jq works by parsing the JSONL format to query specific message roles like user or assistant, and extracting content types such as text from the structured data.

Yes, you need both jq and rg installed to analyze session logs, as these command-line tools are required dependencies to parse and filter the structured JSONL log data efficiently.

Yes, you can calculate session costs from historical conversation logs by applying jq queries to the JSONL files to extract and summarize the cost data associated with past interactions.

You use jq instead of standard grep because session transcripts are structured JSONL files, and jq parses the structured data to filter by message role and content type, whereas rg handles raw text searching.