PromptForge
Prompt engineering workbench for crafting and evaluating LLM prompts.
Description
šš«šØš¦š©šš šØš«š š is an open-source prompt engineering workbench built by developer insaaniManav, with a Go-based API backend. The project gives prompt engineers a workspace to draft prompts, run them against multiple models, and compare outputs without wiring up custom evaluation scripts for every change. It addresses the workflow gap between writing a prompt in a notebook and shipping a tested version to production.
Key Capabilities:
ā Prompt drafting and iteration interface with side-by-side output comparison
ā Systematic prompt evaluation across multiple LLM providers and model versions
ā Prompt versioning with diff tracking between iterations
ā Test case management for regression checks on prompt changes
ā Analysis tools for scoring response quality across runs
ā Self-hostable deployment via Go API backend
ā Open source under a permissive license
Alternative tools
- Langfuse
Open-source LLM engineering platform for tracing, evaluation, and prompt management.