Hello there. I'm new to AI and recently discovered AI coding skills (like OpenCode/Claude Code).
I noticed that AI tends to "hallucinate" - making things up or forgetting context in long sessions. After some research, I learned this happens because the context window has limits.
So I built a simple skill called nhallucinate:
https://github.com/AlanNobita/nhallucinate
It stores memory in project-specific files (.agent/nhallucinate/) so the AI remembers what worked and what didn't across sessions.
Feel free to give me suggestions, teach me new stuffs and correct me what I did wrong here. I am posting this to get a good feedback from people who understand. Thank you!
Top comments (0)