Ggboykxz/APEX — reverse-engineered prompt
Reverse engineered prompt
Build me a polished terminal AI coding agent called APEX. I want it to feel like a serious tool I can install, open in my terminal, connect my API keys, and ask it to edit code, read files, run commands, write tests, and explain what it is doing.
It should support lots of AI providers, including OpenAI, Anthropic, Google, Mistral, DeepSeek, Groq, xAI, and local Ollama models. Let me switch models during a session, track token usage and cost live, resume old sessions, export sessions, and run a one time prompt from the command line.
Please include a nice terminal interface with themes, streaming responses, a command palette, file mentions with @, inline shell commands with !, and keyboard shortcuts. Add different agent modes like coder, architect, planner, shell, reviewer, explorer, and general assistant, with read only modes where appropriate.
Also include setup scripts, config files, docs, tests, Docker support, and sensible security checks for shell commands. Look up current docs online if you need to.
Want more depth? Deep Reverse