session-logs

To search past conversation history in JSONL session logs, use jq and ripgrep-style filtering to locate specific phrases, roles, or tool usage across locally stored transcripts. This extracts matching user and assistant text for immediate recall without manual scrolling.

Summarizing cost and usage data from old agent sessions involves reading JSONL transcript files and computing per-session totals, activity counts, and time ranges. This provides an aggregate view of historical expenses and usage patterns for auditing.

Yes, you can use jq and rg to filter conversation transcripts by participant role. These tools apply jq/rg-friendly patterns to JSONL files, allowing you to narrow results specifically to user or assistant messages for targeted forensic recall.

Finding a specific keyword across multiple parent chat sessions requires cross-session searching across all locally stored JSONL files. Using keyword search patterns, it extracts and returns all matching text occurrences from your historical context.

Yes, you need jq and rg installed to perform conversation forensics on local session files. The Skill depends on these specific command-line tools to parse JSONL structures and apply rapid text filtering for extracting matching messages and aggregates.