oracle

Run long AI queries in persistent tmux sessions with the Oracle CLI.

Updated Jan 2, 2026

Search Tags:#code review #summarization #analysis #tmux #long-running #oracle #AI queries

Ready to run

npx skills add https://github.com/mrud/skills --skill oracle-mrud

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: oracle
Source: https://github.com/mrud/skills/tree/main/oracle
Command: npx skills add https://github.com/mrud/skills --skill oracle-mrud

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill enables running long-running AI queries in a controlled console, preserving session context and progress for review, iteration, and collaboration.

Core Features & Use Cases

Persistent sessions: Start and resume long tasks in a tmux session, ensuring work is not lost.
Code reviews & analysis: Support deep reviews, summarization, and multi-step analysis that require time.
Session-based workflows: Organize tasks with descriptive slugs and maintain reproducible results across steps.

Quick Start

Launch a long-running query in a dedicated tmux session: oracle -p "Summarize the risk register" --slug risk-register-summary --file docs/risk-register.md docs/risk-matrix.md

Frequently Asked Questions about oracle

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

FAQPage Schema

How do I run long AI queries without losing session context if the terminal disconnects?▼

Run long AI queries inside a persistent tmux session to preserve context and progress. This approach keeps tasks like code reviews running safely in the background, allowing you to pause and resume without losing work.

Can I pause and resume a long-running code review or analysis task?▼

Yes, you can pause and resume long-running code reviews or analyses by executing them in a persistent tmux session. This ensures session context and progress are preserved for later iteration and collaboration.

Do I need a tmux environment to manage and execute persistent AI tasks?▼

Yes, a tmux environment is required to manage and execute persistent AI tasks. It provides the controlled console necessary to preserve session context and progress for long-running queries.

What's the best way to organize multi-step analysis workflows with descriptive identifiers?▼

Organize multi-step analysis workflows by launching queries with descriptive slugs. This session-based workflow approach maintains reproducible results and keeps complex tasks structured across steps.

How do I pass multiple files into a summarization task in a persistent session?▼

Pass multiple files into a summarization task by using the file flag followed by file paths. This allows the persistent session to ingest and process multiple documents for deep reviews and summarization.

Why does my long AI query lose progress when I close the console?▼

Long AI queries lose progress when run in a standard console because the session context is destroyed on close. Using a persistent tmux session preserves the state, preventing work from being lost.