unliftedq

I built this tool to resolve my own pain point of agentic automation:

1. Existing CLI solution is provider specific, like minimax cli, chatgpt cli, etc. and for other providers, there's no built-in CLI support.

2. For local CLI/scripts solution, the generation result/history is not tracked, sometimes, I want to generate a similar image, I have to keep the prompts in a notebook. Now, with imagent, I can simply remix any prompt from the library.

3. CLI is the best solution for agent automation, I can use the cli to generate slides, blog illustrations, website assets, etc. And with it, I can even generate videos with hyperframes/remotions with great illustrations and speech audios. All it done by agent, I don't need to create images, audios by myself.

4. Agent isn't aware of the difference/limitation of different models, by maintain the catalogs, agent can discover what is available and choose the best options as its need. And call it in a unified interface.

So, the cost of video generation is not my focus at this point, what I care is automation. If we give the agent such ability, what the agent can create for us automatically. Not vibe code, but vibe creation.

Totally agree, a single one-shot prompt can't prove anything.

I'm tired of big news in this way - a small set of tests to declare one model is better than another, can they really consistently reproduce the result? And there's basically no disclosure: nothing other people can really hand on to verify the tests/judgement by themself.

The best valuable part of DeepSeek V4 pro is its low price, I don't expect have much better performance than GPT-5.5, even it's just the performance like gpt-5.4, it's still a good model.

Good point.

Today, you send such workflow definition to claude code or codex and ask it to delegate proper subagents to complete those tasks in order, it can do quite well. So, what's the unique value for this solution?