session-logs

You can search past conversation logs stored in JSONL files by using the session-logs Skill, which leverages the `jq` and `rg` command-line utilities to query and retrieve historical user and assistant messages.

Yes, you can analyze token costs and tool usage from previous chat sessions by querying the JSONL session log files with `jq` to extract and aggregate the specific cost and tool usage fields.

Yes, you need to install both `jq` and `rg` command-line utilities on your system, as the session-logs Skill directly depends on them to parse and search the JSONL conversation history files.

The best way to extract a specific decision from a past chat session is to use `rg` for targeted text search across JSONL files, and then apply `jq` to filter the exact conversation messages containing the relevant details.

To retrieve all assistant messages from a specific date, you can use the session-logs Skill to filter JSONL records by date and message role, utilizing `jq` to parse the structured conversation history data.

When parsing large JSONL conversation transcripts, processing limitations may arise if system memory is constrained, though combining `jq` for structured parsing with `rg` for fast text searching helps manage large log files efficiently.