Hi everyone 
QoderWork just shipped a major new feature: Awareness. It’s a system that gives your AI assistant the ability to remember, reflect, and grow — three things
that have been missing from AI assistants until now.
Here’s how it works.
The Problem
AI assistants are great at chatting, but they don’t build a relationship with you. Preferences you set today are gone tomorrow. Mistakes you corrected last week come back this week. The product is still a stateless chatbox.
Awareness changes that.
Memory — Your AI Remembers, and You Control It
Memory in QoderWork isn’t just stuffing old conversations into context. It works in layers:
- Short-term memory — daily conversation details
- Long-term memory — things that matter across sessions, promoted from short-term
- Organized in 4 layers — who you are, your work rules, what you’re working on, how you like to be treated
The key detail: All memories are stored as Markdown files on your local machine. You can view, edit, or delete any entry. Nothing is hidden. You have final say over what the AI remembers.
Reflection — Knowing What to Forget
After a session ends, QoderWork can automatically reflect:
- Promote — Move valuable short-term memories into long-term
- Correct — Find outdated or conflicting entries and propose updates
- Rollback — If a reflection crosses a hard rule, it automatically reverts to a pre-reflection snapshot
Think of it as an editor for your notes. Memory without reflection is just a growing pile of noise. With reflection, it becomes a self-organizing notebook.
Skill Evolution — Repeated Work Becomes Reusable Ability
When QoderWork detects that you’ve done the same type of task multiple times, it suggests turning it into a reusable Skill:
- AI generates a skill suggestion with trigger words, templates, and parameters
- You decide — adopt, ignore, or reject
- Rejected suggestions enter cooldown and won’t bother you again
The entire Awareness system adds under 5% overhead to your main conversation. It runs on fork sessions that reuse the original context prefix, keeping prompt cache hit rate above 99%.
How to Enable
Go to Settings → Awareness to turn it on.
Awareness is now live for all QoderWork users. Give it a try and let us know what you think!