suraj-dhal/reverse-prompt-engineer
v1.0.0MIT
Reconstruct one plausible, evidence-led, copy-ready prompt from an existing AI output and its context, preserving requirements, provenance and uncertainty.
plugin.json
- name
- reverse-prompt-engineer
- version
- 1.0.0
- description
- Reconstruct one plausible, evidence-led, copy-ready prompt from an existing AI output and its context, preserving requirements, provenance and uncertainty.
- author
- Suraj Dhalsurajbbsr2007@gmail.comhttps://surajdhal.com
- license
- MIT
- keywords
- prompt-engineeringreverse-promptingprompt-reconstructionprompt-auditcontext-preservationagent-skillsevaluation
- extensions
- com.openai
plugin.json
{
"$schema": "https://agent-plugins.org/schemas/1.0.0/plugin.schema.json",
"name": "reverse-prompt-engineer",
"version": "1.0.0",
"description": "Reconstruct one plausible, evidence-led, copy-ready prompt from an existing AI output and its context, preserving requirements, provenance and uncertainty.",
"author": {
"name": "Suraj Dhal",
"email": "surajbbsr2007@gmail.com",
"url": "https://surajdhal.com"
},
"homepage": "https://github.com/Suraj-Dhal/reverse-prompt-engineer#readme",
"repository": "https://github.com/Suraj-Dhal/reverse-prompt-engineer",
"license": "MIT",
"keywords": [
"prompt-engineering",
"reverse-prompting",
"prompt-reconstruction",
"prompt-audit",
"context-preservation",
"agent-skills",
"evaluation"
],
"extensions": {
"com.openai": {
"interface": {
"displayName": "Reverse Prompt Engineer",
"shortDescription": "Rebuild prompts from outputs",
"longDescription": "Reverse Prompt Engineer reconstructs one plausible, copy-ready user-level prompt from an existing output (a report, article, analysis, code sample, image, website or design) together with any conversation history and files you supply. It keeps material requirements, figures, scenarios and conditional logic; separates explicit evidence from observation and inference; flags conflicts and inaccessible sources; and states an overall confidence level. It can also audit which of your original instructions an output followed, reconstruct one shared prompt for several outputs, or suggest clearly labelled improvements. Intended for writers, analysts, developers, marketers and educators who need to reproduce or document how an AI result was made. Limitations: it cannot recover an exact historical prompt or any hidden system prompt, cannot identify the model or settings used, and does not verify facts or citations unless you ask and a browsing tool is available.",
"developerName": "Suraj Dhal",
"category": "Productivity",
"capabilities": [
"Interactive"
],
"defaultPrompt": [
"What prompt produced this output?",
"Rebuild the master prompt from this chat and files",
"Audit this answer, then reconstruct its prompt"
],
"websiteURL": "https://surajdhal.com",
"brandColor": "#FF4D00",
"composerIcon": "./assets/icon.png",
"logo": "./assets/icon.png"
}
}
}
}
What else this package ships
These files come with the package and this site does not publish them. They are listed so you know what is there before you install it.
- LICENSE
Client extensions
Data this package carries for particular clients. The directory lists the clients named and never reads what is addressed to them.
- com.openai