What is kimi-memory AGENTS.md?

A repository instruction file for the kimi-memory project. It describes the code layout, database storage, hooks, command-line interface, and other parts an agent may need to understand before making changes.

Use it when developing or reviewing kimi-memory, especially changes to memory storage, tools, hooks, the command-line interface, or server wiring.

It gives coding agents a reliable map of the project and its conventions, reducing guesses about where features and data-handling code belong.

How cbuntingde/kimi-memory configures Claude Code and Codex

Reuse

Borrowing it

Nothing to install: this file belongs to cbuntingde/kimi-memory. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/cbuntingde/kimi-memory/main/AGENTS.md

Clone the repo

git clone --depth 1 https://github.com/cbuntingde/kimi-memory

Made for: Claude Code, Codex, OpenCode.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/cbuntingde/kimi-memory/agents-md/github.svg)](https://agentmods.dev/instructions/cbuntingde/kimi-memory/agents-md)

Your own site

<a href="https://agentmods.dev/instructions/cbuntingde/kimi-memory/agents-md"><img src="https://agentmods.dev/badge/instructions/cbuntingde/kimi-memory/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/instructions/cbuntingde/kimi-memory/agents-md"><img src="https://agentmods.dev/badge/instructions/cbuntingde/kimi-memory/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 3,056 This file is loaded in full into every session.

When invoked 3,056 The same file — it is already loaded in full.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.03056	$0.03056
Opus 5.5	$0.01222	$0.01222
Sonnet 5.5	$0.00611	$0.00611
Haiku 4.5	$0.00306	$0.00306

Measured 1mo ago against content hash 10b1636a0d28, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

kimi-memory AGENTS.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

AGENTS.md · 182 lines

How it starts

The opening of the file, as written. The whole thing — 182 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

Orientation for agents working in this repository. README.md covers installation and everyday use; CONTRIBUTING.md covers the change workflow; SECURITY.md covers the threat model. This file is the reference for the code layout and the environment variables.

Code layout

src/mcp/ — the MCP surface. launcher.js is the stdio entry point, main.js builds the server, tool-defs.js holds every tool definition, and handlers/ holds one registration module per domain, wired up with registerTool(server, D.<name>, …).
src/server.js — the server wiring that calls each domain's register().
src/persist/ — the SQLite layer. connection.js owns the schema and the MIGRATIONS array; memories.js, edges.js, share.js, project.js, and the rest hold the row-level helpers.
src/hooks/run.js — the hook dispatcher. Per-event logic lives in src/hooks/handlers/; the top-level hooks/*.js files are thin shims that set KM_HOOK_EVENT and call the dispatcher.
src/cli.js + src/cli-cmd/ — the standalone CLI.
src/proxy/ — the HTTP transport adapter.
src/advisor/ — the advisor keyword detector.
src/<name>.js — feature modules (consolidate, dream, dreaming, auto-gc, extract, embedding, session-focus, work-log, decay, …).
skills/, commands/ — the agent-facing skill and slash-command docs.
tests/NN-<name>.test.js — number-prefixed so they sort in the order they were added.

Environment variables

This is the authoritative table. The README's ## Configuration section covers only the everyday settings; everything the code reads is here. Unless noted, a value of off disables the feature and the default is on.

Core

Variable	Default	What it does
`KIMI_CODE_HOME`	`~/.kimi-code`	Root of Kimi's data folder. Every memory DB and state file lives under `$KIMI_CODE_HOME/kimi-memory/`.
`KIMI_MEMORY_LEGACY_SUBSYSTEMS`	`on`	Set to `off` to skip registration of the 15 legacy MCP tools and the tier/persona sweeps. See "Subsystem deprecation".
`KIMI_MEMORY_SECRET_SCAN`	`on`	Set to `off` to bypass the credential-shape gate on save. Intended for fixture imports only.
`KIMI_MEMORY_BUSY_TIMEOUT_MS`	`30000`	SQLite `busy_timeout` for the connection. A hook process pins this down to 1500 ms (`src/hooks/run.js`) so a write that collides with a concurrent hook fails fast into the handler's fail-open path instead of blocking the JS thread past the dispatcher ceiling — timers cannot run inside a native SQLite call, so a long busy_timeout makes the hook timeout guard unenforceable.

Read the full file on GitHub · 182 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

1mo ago Changed · +1 lines · +57 tokens per session 10b1636a0d28
1mo ago Changed · +1 lines · +132 tokens per session d2ff501f1cfd
1mo ago Changed · -71 lines · -903 tokens per session 59becb194cd5
1mo ago First seen · 251 lines · 3,770 tokens per session scan A d8c60ec3fc75

Subscribe to this mod's changes

kimi-memory AGENTS.md is an instructions file published in the GitHub repository cbuntingde/kimi-memory (0 stars, last pushed 1mo ago), licensed MIT. It adds 3,056 tokens to every session, about $0.0122 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-15.

Building agents

kimi-memory AGENTS.md next to 6 instructions from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
kimi-memory AGENTS.mdcbuntingde/kimi-memory	Instructions file	0	1mo ago	MIT	A	3,056	Claude Code, Codex, OpenCode
vscode buildNext.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	9,719	GitHub Copilot
vscode oss-third-party-notices.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	5,516	GitHub Copilot
next.js AGENTS.mdvercel/next.js	Instructions file	143k	yesterday	MIT	A	7,461	Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchain	Instructions file	149k	today	MIT	A	4,680	Codex, OpenCode
vscode copilot-instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	3,093	GitHub Copilot
vscode model-prompts.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	2,022	GitHub Copilot

kimi-memory: Instructions file for Claude Code