Wild_Root_Prompt logo

Wild_Root_Prompt

Transform any raw prompt into expert-level output using local LLMs via Ollama — no cloud API keys, no data sent to external servers.

100% Local 173 Techniques MIT License macOS · Windows · Linux · Android v2.4 — CI ✓ · CodeQL ✓ · 89 tests

Why Wild_Root_Prompt?

Research-backed techniques

173 prompt engineering techniques across 15 categories — Chain-of-Thought, Tree-of-Thought, ReAct, MECE, Constitutional AI, red teaming, and more — applied before your LLM ever sees the input.

Runs 100% locally

No API calls to external servers, no subscription fees. Powered by free, open-source Ollama. Your data never leaves your machine.

Dual-model synthesis

Run two LLMs in parallel with split-screen streaming, then synthesize both outputs into a single unified, superior document.

Novice & power-user friendly

One-click web UI for beginners, full-featured CLI with scripting support for power users. Compiled single-icon apps for macOS, Windows, and Linux.

Smart technique auto-selection

Describe your task in plain language — the 25-topic index picks the right subset of techniques automatically, in English or French. All 173 techniques reachable, zero configuration.

Privacy & security first

Built-in PII anonymization strips emails, IPs, phone numbers and MAC addresses before prompts leave your machine. SSRF-protected web enrichment, prompt injection detection, CodeQL-scanned codebase.

Prompt Engineering Techniques

173 techniques across 15 categories, including:

See the full list on GitHub →

Quick Start

macOS / Linux:

git clone https://github.com/TFD-42/Wild_Root_Prompt.git && cd Wild_Root_Prompt && chmod +x install.sh && ./install.sh

Or download a compiled app — no Python required:

Related Resources