Kimi Code plugin: three-layer local memory (global user, per-project durable + working, per-project session archives) plus an integrated advisor subsystem for reflective prompts. Exposed via MCP tools, lifecycle hooks, and slash commands.
This repository also configures its own agents. See what kimi-memory tells them →
Skill Claude CodeCodex
Reflect on the active project and propose what to change, do differently, or consider that has not been considered. Invoke when the user asks "would we change anything", "what would you do differently", "how can we improve", "what are we missing", "is there a better way", "second opinion", "review this approach"…
Skill Claude Code
Persistent memory for Kimi Code across three layers (global user preferences, project decisions, session archives). Use when the user asks to remember, recall, or persist a fact, preference, decision, or convention. Also use at session start to recall context.
Skill Claude CodeCodex
Show the kimi-memory entries currently on file for this project (and the global layer when relevant). Use when the user asks to "list memories", "what do you remember", "show project memory", or "listmemories".
Skill Claude Code
Wipe the per-project memory cache for the active project when it was re-cloned from the same path. Use when the user says "reset project memory", "wipe this project's memory", "I re-cloned the repo", or "memory is stale".
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: