MemTensor/memmy-agent

C3

🍙 A personal AI agent & local memory hub for all AI agents, gives every AI one shared, fully controlled memory and persistent context — all AI remember the same you. Now supports Claude Code, Codex,

★ 2.0k · MemTensor/memmy-agent source on GitHub · this plugin in the registry

MemTensor/memmy-agent is a DeepSeek Harness plugin rated C3 — powerful capability combined with sensitive behavior. It runs code at install time, can rewrite the system prompt, executes system commands, decodes base64 payloads.

Not a plugin — depends on dsh packages but declares no plugin manifest

What it can do

CapabilityFlagEvidence
runs code at install timeinstall_scriptpostinstall: node scripts/postinstall.js
can rewrite the system promptprompt_surfacesystemPrompt
executes system commandsexec×8 in authored code, e.g. App/backend/src/infrastructure/app-state-store/local-data-store.ts:2, App/backend/src/infrastructure/app-state-store/local-data-store.ts:122
decodes base64 payloadsbase64_decode×3 in authored code, e.g. App/backend/src/infrastructure/app-state-store/secret-store.ts:68, App/backend/src/infrastructure/app-state-store/secret-store.ts:69
reads credential-class env varstoken_envMEMMY_SECRET_KEY
ships no manifestno_manifestuses 5 services, 9 tool regs, no manifest

Services it injects

agents sessions systemPrompt tools uiConversation

Hooks it attaches

agent/pre-step session/event

Outbound domains

dashscope.aliyuncs.com api.anthropic.com api.openai.com www.googletagmanager.com discord.gg x.com open.dingtalk.com open.feishu.cn docs.discord.com core.telegram.org

Environment variables it reads

MEMMY_CONFIG MEMMY_APP_DB_PATH MEMMY_CLOUD_SERVICE APPDATA XDG_DATA_HOME MEMMY_SECRET_KEY MEMMY_APP_EDITION OPENCLAW_WORKSPACE_DIR OPENCLAW_PROFILE

How to read this

Levels measure capability surface and transparency, not maliciousness. A C3 plugin can be entirely legitimate — a desktop shell genuinely needs subprocesses. The point is that you can see this before installing. See the levels explained and how dsh plugins work.

Findings come from static analysis of shipped code; nothing is executed. Think a flag is wrong? Open an issue — every flag cites the file and line it came from.