How songkai-liu-sjtu/idesigngpt works
iDesignGPT is a prompt-based engineering-design workflow intended to run in FastGPT, not a conventional application with an in-repository runtime. Users load design-method prompts into FastGPT to support problem refinement, information gathering, design-space exploration, and iterative optimization. The repository also contains independent quantitative-analysis scripts and evaluation materials. Prompt files are represented as distinct design-method capabilities; the README does not specify their execution order or the internal wiring of FastGPT workflows, so those relationships are not inferred.
Main entry points: Brainstorming.md Brainstorming · Functional Structure Decomposition.md Functional decomposition · Morphological Matrix.md Morphological analysis · Scamper.md SCAMPER
Components 19
Design prompts
- Brainstorming
code/agent/Brainstorming/Brainstorming.md - Functional decomposition
code/agent/Functional Structure Decomposition/Functional Structure Decomposition.md - Morphological analysis
code/agent/Morphological analysis/Morphological Matrix.md - SCAMPER
code/agent/Scamper/Scamper.md - TRIZ
code/agent/Triz/Triz.md - Biomimetic design
code/agent/Biomimetic design/Biomimetic Design.md
Case-study analysis
- Search-query wordcloud
code/quantitative insights into case study/fig4b_wordcloud.py - Topic modeling
code/quantitative insights into case study/fig4cd_topic_modeling.py - Design-space analysis
code/quantitative insights into case study/fig4ef_design_space_analysis.py - Coverage and novelty
code/quantitative insights into case study/fig4f_coverage_diversity_vovelty_analysis.py
Evaluation
- Delphi evaluation
code/performance evaluation - Questionnaire analysis
code/user experience/questionnaire_analysis.m - Objective measures
data/Performance_evaluation/Objective_measures
Study data
- User evaluation
data/User_Evaluation - Case-study results
data/Case_study - Generated design outputs
data/Raw_output - Questionnaire responses
data/Questionnaire
Other components
- Design engineer
- FastGPT platform
Connections 7
- Design engineer → FastGPT platform: runs workflows
README.md - FastGPT platform → Brainstorming: runs prompt
README.md - FastGPT platform → Functional decomposition: runs prompt
README.md - FastGPT platform → Morphological analysis: runs prompt
README.md - FastGPT platform → SCAMPER: runs prompt
README.md - FastGPT platform → TRIZ: runs prompt
README.md - FastGPT platform → Biomimetic design: runs prompt
README.md