Summary
Start exploring a mock-level GUI frontend for multi-ai-cli.
This is not intended to be a full production GUI yet.
The goal is to experiment with how existing CLI workflows can be translated into a graphical interface and to validate whether a lightweight desktop frontend is useful.
Background
multi-ai-cli already provides a rich command model, including:
- interactive REPL usage
- sequence-based execution
- file-based input/output workflows
- shared
work_data style artifact handling
While the CLI remains the primary interface, a small GUI prototype may help:
- users who are less comfortable with terminal usage
- visual editing of flow definitions
- browsing prompt / output files
- basic REPL-style interaction through a persistent backend process
During initial prototyping, it became clear that the GUI behaves largely like a visual frontend for existing @sequence-style workflows, which is a useful observation in itself.
Goals
- create a mock-level GUI frontend for experimentation
- validate usability of flow editing and file browsing in a desktop UI
- test simple REPL integration via
multi-ai-cli --mode repl
- observe which parts of the current CLI map naturally to GUI interaction
Non-Goals
The following are out of scope for now:
- production-ready GUI architecture
- full feature parity with the CLI
- structured machine-readable backend protocol
- complete separation of GUI/core/adapter packages
- polished UX, packaging, or release workflow
Current Direction
The initial GUI prototype is expected to include:
- a flow / sequence editor
- a terminal/log display
- prompt and output file browsing
- simple interaction with a persistent
multi-ai-cli --mode repl backend
At this stage, the GUI should be treated as experimental and mock-level.
Expected Outcome
This issue is mainly for exploration.
Success means:
- the GUI can be used to test interaction ideas
- we can identify what works well as a GUI
- we can clarify whether the GUI is mainly:
- a sequence/flow workbench
- a REPL frontend
- or a combination of both
This issue should help guide future architectural decisions without forcing premature redesign.
Summary
Start exploring a mock-level GUI frontend for
multi-ai-cli.This is not intended to be a full production GUI yet.
The goal is to experiment with how existing CLI workflows can be translated into a graphical interface and to validate whether a lightweight desktop frontend is useful.
Background
multi-ai-clialready provides a rich command model, including:work_datastyle artifact handlingWhile the CLI remains the primary interface, a small GUI prototype may help:
During initial prototyping, it became clear that the GUI behaves largely like a visual frontend for existing
@sequence-style workflows, which is a useful observation in itself.Goals
multi-ai-cli --mode replNon-Goals
The following are out of scope for now:
Current Direction
The initial GUI prototype is expected to include:
multi-ai-cli --mode replbackendAt this stage, the GUI should be treated as experimental and mock-level.
Expected Outcome
This issue is mainly for exploration.
Success means:
This issue should help guide future architectural decisions without forcing premature redesign.