#117117th of 149 in Skills, plugins and rules
oracle
CLI and MCP server that sends a prompt plus selected files to a second model
Documentation ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 4.1k
- License
- MIT
- Last commit
- Oct 2026
- Last release
- Oct 2026
- Language
- TypeScript
Missed the latest GitHub check; these facts may be older.
Overview
Oracle bundles a prompt with files you select (globs, directories, exclusions) and sends it to a model through a provider API or a signed-in ChatGPT or Gemini browser session. Results are stored as sessions under ~/.oracle/sessions and can be reattached, replayed or followed up. It also runs as an oracle-mcp stdio server and ships a skill so coding agents can ask for a second-model review.
Who it is for: Developers and coding agents wanting a second-model review of project files
Strengths
- Render mode prints the exact prompt and files without credentials or a model call
- API mode supports OpenAI, Azure OpenAI, Anthropic, Gemini, xAI, OpenRouter and compatible endpoints
- --models runs several models in one session with per-model usage, cost and partial failures
- Dry run reports resolved files and token estimate; bundles carry line numbers for path:line citations
Weaknesses
- Requires Node.js 24 or newer
- Browser mode depends on Chrome automation and a one-time login flow
- Follow-ups work only for supported API or ChatGPT conversations
- No Docker image or demo listed in the README
What it needs
- no GPU
- Needs Node.js 24+, Chrome (browser mode), Provider API key (API mode)
- Models: OpenAI, Azure OpenAI, Anthropic, Gemini, xAI
Also in Skills, plugins and rules
See all 149| Rank | Project | Score |
|---|---|---|
| 1 | cavemanCuts token use in coding agents with terse replies and compressed inputs | 95 out of 100 |
| 2 | skillsAgent skills for planning, TDD, debugging and code review | 94 out of 100 |
| 3 | eccAgents, skills, hooks and rules bundle for Claude Code, Codex and other harnesses | 94 out of 100 |
| #4 | impeccableDesign skill with 24 commands and a detector for AI-generated frontends | 94 out of 100 |
| #5 | ponytailSkill prompt that makes coding agents write less code and add tests | 92 out of 100 |