skills.rest
MOODMNKY-LLC
MOODMNKY-LLCCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

oracle

Run targeted code review tasks with token-efficient dry-run previews.

Updated May 7, 2026
Search Tags:#code review#debug#refactor#oracle cli#token preview#browser engine#second-model analysis
One-click install
npx skills add https://github.com/MOODMNKY-LLC/mnky-atlas --skill oracle-moodmnky-llc
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: oracle
Source: https://github.com/MOODMNKY-LLC/mnky-atlas/tree/main/skills/oracle
Command: npx skills add https://github.com/MOODMNKY-LLC/mnky-atlas --skill oracle-moodmnky-llc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Software engineers often waste time manually reviewing, debugging, or refactoring code across large file sets, and struggle to get actionable, context-aware feedback from secondary AI models without loading irrelevant files or overspending on tokens.

Core Features & Use Cases

  • Targeted Second-Model Analysis: Run review, debug, refactor, or design tasks on a curated set of project files, avoiding irrelevant context that dilutes model output.
  • Token-Efficient Previews: Use dry-run modes to validate payload content and estimate token spend before running full analyses, preventing unexpected costs.
  • Flexible Execution Engines: Choose between browser mode for long-running Pro model runs or API mode for fast, explicit calls with OpenAI, Claude, or other supported models.
  • Use Case: A frontend engineer can use this skill to get a focused refactor plan for their React component library by passing only the relevant component files, previewing token costs first, then running the analysis in browser mode for deep, long-form feedback.

Quick Start

Use the oracle skill to preview the token cost and payload for a code review task targeting all non-test files in your project's src directory before running a full analysis.

Frequently Asked Questions about oracle

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

FAQPage Schema
How do I run a second-model code review on specific project files?▼

You can perform a second-model code review by passing a curated set of project files to the analysis engine. Using glob-based inclusion and exclusion rules prevents loading irrelevant context, ensuring the model provides targeted feedback without diluting the output.

How do I estimate token costs before running an AI code analysis?▼

You estimate token costs for AI code analysis by using a dry-run mode to validate the payload content. This token-efficient preview calculates expected token spend and verifies file selection before executing the full model run, preventing unexpected costs.

Can I use a browser engine for long-running AI model code refactoring?▼

Yes, you can use a browser execution engine for long-running AI model code refactoring. Flexible execution engine selection supports browser mode for sustained Pro model runs, while API mode handles fast, explicit calls with models like OpenAI or Claude.

What is the best way to prevent secret leakage during AI code debugging?▼

The best way to prevent secret leakage during AI code debugging is through safe context handling and targeted file selection. By curating project file sets with glob-based exclusion rules, you filter sensitive data out before generating the analysis payload.

Does this approach support session management for long-running model runs?▼

Yes, this approach supports session management specifically for long-running model runs. This functionality maintains state during deep code refactoring or design feedback operations, ensuring continuous analysis when using browser execution mode.

Related Skills

Explore More Software Engineering Skills→
hereinthehivehereinthehive
5

test-assumption

Analyze code to surface hidden user assumptions and generate a structured inclusion report.

Community
Intermediate
jamespacileojamespacileo

fix-application

Apply codebase fixes through a phase-guided Kaizen workflow.

Community
Advanced
julianobarbosajulianobarbosa
7

reviewing-code

Review code for quality, bugs, security issues, and best practices using Codex.

Community
Advanced
bocanbocan
27

localization-setup

Generate type-safe localization infrastructure for iOS and macOS apps.

Community
Intermediate