RECALL · CONCERN · SELF-AUDIT
kimi-core
@MARIKAGURA
puts memory, concern, emotional drive, and active wake-up into a set of auditable 1v1 kernels.
kimi-core uses PostgreSQL/pgvector to save events, states, entities and images, and combines vectors, keywords, time and importance into hybrid searches. The active layer determines actions based on drive and concern, and also allows the option of doing nothing.
Project address (can be copied to AI):https://github.com/marikagura/kimi-core
PROJECT INTRO
Project introduction
kimi-core is a set of memory and autonomy cores for single humans and single AIs. It provides vector, BM25/fuzzy keyword, time decay and importance fusion retrieval, and uses append-only events and manual curation to prevent automatic summarization from silently corrupting memory.
Self-drive, concern, scheduled wake, reentry/closeout session life cycle and reproducible retrieval evaluation are all connected; external actions use human-in-the-loop by default. The project does not come with a fixed personality, and the relationship layer is established by the user himself.
- Suitable for:Heavy 1v1 Agent users who require auditable memory, active concern, and rigorous retrieval evaluation.
- Platform:Node.js, PostgreSQL/pgvector and HTTP MCP; active daemon optional.
- Usage:Initialize personality and configuration, start Gateway; run daemon when active wake-up is required.
- License:AGPL-3.0-or-later。
COMMENTS & FEEDBACK
Comments and Feedback
People who have used this project can come back and tell newcomers: which platform it ran on, whether there were any pitfalls during the installation, and what the actual experience was like. The current version is in guest mode, and you do not need to register an account to leave a message.
The message is waiting for review and will be publicly displayed here after approval.
SOURCE & CREDIT
Atlas is responsible for the introduction and navigation, and the use still returns to the original author.
We do not mirror project files, nor do we intercept author traffic. For installation, download, version updates and the latest instructions, please refer to the project page provided by the original author.
Reading public comments...