skills.rest
tommy0103
tommy0103COMMUNITY
Explore all skills by this author →
GitHub

obelisk

Query Claude Code and Codex session history via JavaScript against a SQLite index.

305|19|Updated May 29, 2026
Search Tags:#indexing#search#query#session history#memory layer
One-click install
npx skills add https://github.com/tommy0103/obelisk --skill obelisk
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: obelisk
Source: https://github.com/tommy0103/obelisk/tree/main
Command: npx skills add https://github.com/tommy0103/obelisk --skill obelisk

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill provides a way to search and query Claude Code and Codex session history, enabling users to easily retrieve past session information and make informed decisions based on past interactions.

Core Features & Use Cases

  • Search and Query: Allows users to search and query their Claude Code and Codex session history.
  • Session History: Stores and indexes Claude Code and Codex session history.
  • Agent-First Retrieval: Provides an agent-first approach to retrieving session history, allowing agents to write JS queries and run them locally.
  • Memory Layer: Supports a persistent memory layer for storing and recalling conclusions from past sessions.
  • App Surface: Provides an Electron desktop app for humans to browse sessions, manage memories, and view usage stats.

Quick Start

To install the skill, run the following command:

npx skills add tommy0103/obelisk-skill

Then, in any Claude Code session, you can use the skill by running:

/obelisk-skill <your question>

Frequently Asked Questions about obelisk

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

FAQPage Schema
How do I search and query Claude Code session history?▼

You can search and query Claude Code session history by indexing transcripts into a SQLite database and running JavaScript queries locally to retrieve past session information and make informed decisions.

What is a memory layer for coding agents and how does it work?▼

A memory layer for coding agents is a persistent storage mechanism that stores and recalls conclusions from past sessions. It works by indexing session transcripts so agents can retrieve historical interactions and inform future decisions.

How do I retrieve past Codex session transcripts for an agent?▼

To retrieve past Codex session transcripts, you index them into a SQLite database and use an agent-first approach to write and run JavaScript queries locally, enabling the agent to access past interactions directly.

Can I browse Claude Code sessions and manage memories in a desktop app?▼

Yes, an Electron desktop app is provided for humans to browse indexed Claude Code sessions, manage the persistent memory layer, and view usage stats without writing JavaScript queries manually.

Do I need to write JavaScript queries to search Claude Code session history?▼

JavaScript queries against a SQLite index are required for the agent-first retrieval approach, but human users can also browse sessions and manage memories directly through the provided Electron desktop app interface.

Related Skills

Explore More Software Engineering Skills→
family3253family3253

add-skill

Install agent skills from GitHub repositories into supported runtimes via the add-skill CLI.

Community
Intermediate
0xBigBoss0xBigBoss
53

extract-transcripts

Extract Claude Code and Codex session JSONL transcripts into Markdown.

Community
Basic
dejanrdejanr
15

improve-skill

Analyze coding agent session transcripts to improve or create skills.

Community
Intermediate
heimeshoffheimeshoff
1

fsharp-persistence

Implement SQLite and JSON file persistence with async CRUD in F#.

Community
Advanced