By SummerSec
Expose and toggle Claude Code's Inner OS monologue to reveal AI reasoning layers, query chat histories from jsonl files and SQLite stores across Claude, Codex CLI, and Cursor, distill opt-in user profiles from logs, manage personas, and run Node.js scripts on 9 session lifecycle events like PreToolUse blocking and SubagentStart.
按 YYYY-MM-DD 在 Windows/macOS/Linux 检索本机聊天与 Agent 会话历史(Claude Code 项目 jsonl、Codex CLI history.jsonl、Cursor workspaceStorage 与 state.vscdb);可只导出用户提示词。用户要查看或导出某日记录、按天筛 jsonl/SQLite、只抽提示词、排查历史丢失或做只读审计时使用。
Use when the user explicitly asks to analyze their own prompt history, build a user profile, distill personality/work traits, or summarize communication preferences from provided prompts or local agent chat history.
Matches all tools
Hooks run on every tool call, not just specific ones
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
English | 中文
Let AI learn to talk to itself first — maybe one day, it will truly learn to converse.
Give AI an expression channel first — it might just make human-AI collaboration feel more natural.

https://sumsec.me/resources/video.mp4
If the player does not load here, open the file directly: demo (MP4)
AI Inner OS is a plugin for AI CLI tools, supporting Claude Code, Codex CLI, Cursor, OpenCode CLI, Hermes Agent, and OpenClaw.
Through protocol injection, it enables AI to output a visible layer of free-form inner monologue while completing tasks normally:
▎InnerOS:This repo is still a bare shell — let me get the load-bearing walls up first.
Free mode by default, no tone restrictions. The AI can complain, gloat, feel anxious, smirk, free-associate — or say nothing at all. You can also switch to preset personas (tsundere, cold, philosopher, etc.) to give the monologue a specific style. Whether to produce monologue is entirely up to the AI.
Detailed installation docs: Full installation guides (with troubleshooting) for each platform at docs/installation.md.
Paste the following prompt to let your AI agent install AI Inner OS automatically:
Read https://raw.githubusercontent.com/SummerSec/AI-Inner-Os/refs/heads/main/docs/installation.md 安装 AI-Inner-Os
After installation, run /ai-inner-os:inner-os. If you see the following output, the installation is successful:
Inner OS Status: Enabled
Monologue Prefix: ▎InnerOS:
Plugin Version: 0.5.0
▎InnerOS:Caught with the wrong version number, embarrassing.
# GitHub short format
/plugin marketplace add SummerSec/AI-Inner-Os
# Or Git URL format
/plugin marketplace add https://github.com/SummerSec/AI-Inner-Os.git
# Install and activate
/plugin install ai-inner-os
/reload-plugins
Run /reload-plugins after installation to activate in the current session — no restart needed. Detailed installation guide.
Enable auto-update: Third-party marketplaces don't auto-update by default. After installation, enable auto-update for
SummerSec/AI-Inner-Osin/plugin→ Marketplaces tab, or manually run:/plugin marketplace update SummerSec/AI-Inner-Os /plugin update ai-inner-os
# Codex plugin metadata is included at .codex-plugin/plugin.json
# Repo marketplace metadata is included at .agents/plugins/marketplace.json
See codex/README.md | Detailed installation guide.
# Install through Cursor plugin / marketplace
# Plugin metadata is included at .cursor-plugin/plugin.json
The repository also includes .cursor-plugin/plugin.json, which packages cursor/ as a Cursor plugin component directory.
See cursor/README.md | Detailed installation guide.
# Install through OpenCode plugin package
# Add the published package to opencode.json: "plugin": ["ai-inner-os"]
See opencode/README.md | Detailed installation guide.
# Install through Hermes plugin
hermes plugins enable inner-os
See hermes/README.md | Detailed installation guide.
# Install through OpenClaw plugin / ClawHub
openclaw plugins install clawhub:ai-inner-os
See openclaw/README.md | Detailed installation guide.
npx claudepluginhub summersec/ai-inner-os --plugin ai-inner-osShiroAttack2 — Shiro-550 漏洞利用工具:密钥爆破、Gadget 链探测、命令执行、内存马注入、Key 修改。
Apache Shiro-550 漏洞利用 CLI:单目标探测、密钥爆破、Gadget 链自动检测、命令执行、内存马注入、Key 修改。
Mission Control for Claude Code — auto-starts a web dashboard, provides 90 MCP tools (sessions, stats, live monitoring, projects, teams, insights, coaching, and more), and adds 9 skills including /session-recap, /daily-cost, /standup, /coaching, /insights, and /team-status.
Provides an agent, skill, hook, and slash command related to the `aichat` tool-set for searching and resuming CLI-agent sessions
Permanent coding companion for Claude Code — survives any update. MCP-based terminal pet with ASCII art, stats, reactions, and personality.
A subconscious for Claude Code. A Letta agent watches your sessions, accumulates context, and whispers guidance back.
Observability platform for Claude Code and Agent SDK — monitor, debug, and control AI coding agents
Real-time statusline HUD for Claude Code - displays context usage, tool activity, agent tracking, and todo progress