session-logs

Extract and filter JSONL session transcripts with jq and rg.

Updated Apr 8, 2026
One-click install
npx skills add https://github.com/Jnuso/OpenClaw --skill session-logs-jnuso
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: session-logs
Source: https://github.com/Jnuso/OpenClaw/tree/main/skills/session-logs
Command: npx skills add https://github.com/Jnuso/OpenClaw --skill session-logs-jnuso

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill helps you locate and analyze historical conversation data stored as JSONL transcripts, so you can recall context, audit interactions, and understand past assistant behavior without relying on memory.

Core Features & Use Cases

  • Query per-session transcripts with jq to extract messages, timestamps, and roles.
  • Filter results with rg to surface relevant interactions and tool results.
  • Audit past conversations to support context reuse, compliance, or debugging.

Quick Start

Search your past sessions by keyword or date to retrieve relevant transcripts and context.

Frequently Asked Questions about session-logs

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I search past JSONL conversation logs for specific keywords?

Search past JSONL conversation logs by applying rg to filter transcripts and surface relevant user messages, assistant replies, and tool results matching your keywords. This extracts prior context quickly without relying on memory.

Can I extract specific message fields from JSONL transcripts using jq?

Yes, you can extract specific message fields from JSONL transcripts using jq to query per-session data and pull out messages, timestamps, and roles. This enables precise extraction and filtering of historical conversation data.

What is the best way to audit past assistant behavior in session logs?

Auditing past assistant behavior in session logs involves locating and analyzing historical JSONL transcripts stored under the state directory. Querying these transcripts reconstructs prior context and reviews interactions for compliance or debugging.

Does this approach work for reconstructing context from varying transcript structures?

Yes, reconstructing context from varying transcript structures works by locating historical session data from JSONL transcripts. The process handles varying structures to surface relevant user messages, assistant replies, and tool results.

How do I filter historical tool results from agent state directories?

Filter historical tool results from agent state directories by applying rg across per-session JSONL logs. This surfaces relevant interactions and tool results stored within the state directory for a given agent.