session-logs

You can search past conversation history by scanning saved JSONL session transcripts. The tool uses ripgrep for fast text matching to locate specific keywords across user and assistant messages stored locally.

Extract older messages from JSONL session logs by parsing files with jq. It filters structured transcript data to retrieve parent conversation context and specific content from sessions saved locally.

Yes, you can calculate token usage cost and message counts from saved session transcripts. The tool uses jq to compute session-level stats, aggregating totals per session or per day.

Summarize tool calls from previous agent sessions by parsing the JSONL transcripts with jq. It inspects session structure to identify which tools were invoked and understand prior actions.

Yes, you need both jq and rg installed. The Skill requires jq for local JSONL parsing and rg for fast text matching while reading files from the agents sessions directory.

The best way to find older context across multiple transcripts is using keyword extraction with rg and jq. It reads saved JSONL files to locate matching user and assistant messages across many sessions.