oracle

Bundle prompts and selected files into a single contextual request.

1|Updated Jan 28, 2026
One-click install
npx skills add https://github.com/Zentin-L/Masterbot --skill oracle-zentin-l
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: oracle
Source: https://github.com/Zentin-L/Masterbot/tree/main/skills/oracle
Command: npx skills add https://github.com/Zentin-L/Masterbot --skill oracle-zentin-l

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Oracle bundles your prompt + selected files into one “one-shot” request so a model can answer with repository context (API or browser automation). Treat output as advisory: verify against code + tests.

Core Features & Use Cases

  • Bundle prompts with a selected file set to create a contextual, code-aware request for models.
  • Attach files via --file patterns, with support for inclusion/exclusion and size limits.
  • Switch between API and browser execution modes and reuse stored sessions for long-running tasks.

Quick Start

Bundle your project files with a task description using oracle to obtain a contextual model answer.

Frequently Asked Questions about oracle

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

FAQPage Schema
How do I bundle multiple project files into a single prompt for code review?

Bundle selected files into a single prompt by using --file patterns to include specific code, allowing the model to provide repository-aware answers. This consolidates your code context into one request.

What is prompt bundling for contextual model answers?

Prompt bundling combines your text prompt and selected project files into one request so a model can respond with repository-aware answers. It supports file inclusion, exclusion, and size limits.

Do I need to install an npm package to bundle prompts with file context?

Yes, you need the oracle CLI installed via the npm package @steipete/oracle to bundle prompts with file context. The Skill uses this binary to package your files and prompt into a contextual request.

Can I use API or browser execution modes for bundled code prompts?

Yes, you can choose between API and browser execution modes when running bundled code prompts. This allows you to reuse stored sessions for long-running tasks depending on your environment and model.

How do I exclude irrelevant files when bundling context for a model request?

You can exclude irrelevant files when bundling context by using specific --file patterns and size limits. This ensures only the necessary code is included in the final contextual request sent to the model.

Should I verify code review answers generated from bundled prompts?

Yes, you should treat output from bundled prompts as advisory and verify it against your code and tests. The model provides repository-aware answers based on the bundled context, but verification is essential.