EricLens

If you use AI dev tools daily, you ve probably felt this:

You start a new session and immediately have to re-explain:

• what the project is

• what you already tried

• why certain decisions exist

• what not to repeat

Not because the AI is bad.
Because the workflow forgets.

Hi community!

At Jots we believe that reflection should be a core engineering skill, especially in this age of AI.

So we built a Reflection Skills MCP and made it open-source.
It is a protocol server that exposes:

Hi Everyone! Our app is designed to be used at site and I always felt that response speed is critical for seamless app use

It appeared to be a compromise. This specially relates to audio and TTS we use. We can use very naturally sounding TTS models, but they may take about 1 minute to respond. Or we can use something which responds in 5 seconds but quality will be lower. How much do you think a user can wait for response before she shuts down the app as it takes too long?