Observed arrival · 2026-09-19
CortHeXis Wants Your AI to Remember the Work
An open-source memory method that turns notes, decisions, and corrections into searchable context for AI agents.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- People building persistent AI workflows from their own files
- Worth noticing
- The method uses one Markdown file per fact, semantic retrieval, MCP, and a check for silently missing memory.
Field notes
The workflow stores facts as individual Markdown notes, adds descriptions and keywords, and retrieves relevant material at the beginning of an agent task. Corrections can be encoded as rules with their reasons, while an MCP-based safeguard is intended to catch sessions that start without the memory layer. The page attributes the method to a private operating corpus containing 289 notes, 260,344 words, and 886 commits.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
One card from the complete issue