Why your AI agents keep starting from zero
The problem is not only memory: without a contract, verification and closure, every session explores again and can declare victory without evidence.
Read the investigationJournal / operational evidence
We do not publish hype.
We publish the mechanism.
Research, guides and clear limits for working with AI agents on real repositories.
01
Latest signal
02
Recent stream
How to share context across Claude Code, Codex, Cursor and OpenCode
A practical architecture for letting several agents read common rules without mixing personal memory, team judgment or execution evidence.
Read the investigation
Cardumen on vs. off: what we measure
How we design on/off comparisons with the same repo, model and task; what signals we observe and why the sample still demands caution.
Read the investigation