oracle
Ask the oracle when you're stuck. Invoke GPT-5 Pro with a custom context and files.
Oracle is an MCP server and CLI tool that bundles project files and custom prompts, then sends them to AI models for expert review and analysis. It allows developers and coding agents to invoke a second AI model—such as GPT-4, Claude, or Gemini—with full project context for code audits, security reviews, and architectural feedback. Oracle supports both API-based providers (OpenAI, Anthropic, Azure, xAI, OpenRouter) and browser automation modes for ChatGPT and Gemini web sessions. Sessions are stored locally, enabling background processing, conversation follow-ups, and multi-model comparisons in a single run.
Key Features
Use Cases
- 01Request a second-model code review of pull requests or architectural changes before merging
- 02Audit security-sensitive modules for race conditions, vulnerabilities, or compliance issues
- 03Run multi-model evaluations to compare responses from GPT-4, Claude, and Gemini on the same codebase
- 04Integrate Oracle as an MCP skill in AI coding agents for grounded, context-aware analysis
- 05Generate release risk assessments by bundling package metadata and changelogs for AI review
- 06Preview and export prompt bundles for manual pasting into AI chat interfaces
Related MCP Servers
View moremarkitdown
Python tool for converting files and office documents to Markdown.
firecrawl
The context API to search, scrape, and interact with the web at scale. 🔥
prompts.chat
f.k.a. Awesome ChatGPT Prompts. Share, discover, and collect prompts from the community. Free and open source — self-host for your organization with complete privacy.
langflow
Langflow is a powerful tool for building and deploying AI-powered agents and workflows.
oracle — FAQ
What is the Oracle MCP server?+
Oracle is an MCP server and CLI that bundles project files with custom prompts and sends them to AI models (OpenAI, Anthropic, Gemini, etc.) for code reviews and analysis. It integrates with MCP clients like Claude Code, Codex, and Cursor to provide agents with second-model expertise grounded in actual project context.
How do I install Oracle as an MCP server?+
Install Oracle globally via Homebrew (brew install steipete/tap/oracle) or npm (npm install -g @steipete/oracle), then configure your MCP client to invoke the oracle-mcp stdio server. Node.js 24 or newer is required.
Which AI clients work with Oracle?+
Oracle's MCP server works with Claude Code, Codex, Cursor, and other MCP-compatible clients. The CLI can also invoke API providers (OpenAI, Anthropic, Azure, Gemini, xAI, OpenRouter) or automate ChatGPT and Gemini browser sessions.
Do I need API keys to use Oracle?+
API keys are required for API mode with providers like OpenAI or Anthropic. Browser mode uses signed-in ChatGPT or Gemini sessions via Chrome automation and does not require API keys. Render mode works entirely offline with no credentials.
Is Oracle free to use?+
Oracle itself is open-source under the MIT license. API mode incurs costs from your chosen AI provider; browser mode uses your existing ChatGPT or Gemini subscription. Render mode has no external costs.
Can Oracle run multiple AI models at once?+
Yes, the --models flag runs a multi-model panel in API mode, sending the same prompt to multiple providers simultaneously and recording per-model usage, cost, and output in one session.
How do I install oracle?+
Open the source repository on GitHub and follow its README. oracle is a mcp server — MCP Agents Market links you directly to the official repo.
Is oracle free?+
oracle is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.