We know giving feedback to agents in terminals can be awkward.. large scrollbacks, comments get lost in transcripts, comment on sub-bullets of list nested findings, etc.
With `mdtation` (aka meditation) I try to calm down this flow and make interactions with agents more straightforward.
If you like terminal based tools, herdr/hunk/agentation, I think you'll like this. First, let the agent create a file, open in mdtation to leave comments anywhere, and the agent can follow along.
With `mdtation` (aka meditation) I try to calm down this flow and make interactions with agents more straightforward.
If you like terminal based tools, herdr/hunk/agentation, I think you'll like this. First, let the agent create a file, open in mdtation to leave comments anywhere, and the agent can follow along.
https://mdtation.com