ng:prompt-extractor

Reconstructing a prompt sequence involves mapping original requirements to AI-generated artifacts to produce a reusable, ordered list of prompts. The extractor analyzes the project root and key directories to classify content and determine dependencies among the prompts used.

A vibe-coded project is built using AI-generated code and artifacts. Prompt extraction works by comparing original human-written requirements against these AI outputs to identify gaps, map dependencies, and produce a structured prompt plan suitable for reuse and handoff.

You reconstruct the prompt chain by feeding the vibe-coded project into the extractor. It analyzes the project structure, classifies content, determines prompt dependencies, and outputs a structured sequence detailing the phase, prompt text, input context, and expected output for each step.

Yes, you can extract a structured prompt plan from existing AI-generated code. The tool analyzes your project directories to produce an ordered prompt sequence that includes the phase, prompt text, input context, expected output, and a final report for handoff.

The reconstructed prompt sequence outputs a structured format that includes the phase, prompt text, input context, and expected output for each step. It also generates a final report detailing dependencies among prompts, making it suitable for project handoff and reuse.

Yes, the prompt extractor requires original human-written requirements to compare against AI-generated docs and code. This comparison is necessary to accurately identify gaps, classify content, determine dependencies among prompts, and reconstruct the end-to-end prompt chain.