Security & AI safety

Privacy tools

Privacy tools keep private data out of AI systems or strip identifying traces from content. Pasteguard, for example, acts as a local proxy that redacts sensitive details before they reach a chat model, while remove-ai-watermarks deletes provenance metadata from files you own. The focus is on concrete filtering and anonymization at the point where data leaves your control.

272 agents · most starred first

Top 100 of 272

netbird★ 31,163

NetBird securely connects people, machines and AI agents across any network with one identity-based Zero Trust platform. Powered by peer-to-peer WireGuard®

watermarks-remover★ 27,872

A privacy-first app that strips AI watermarks from content you own.

RemoveWindowsAI★ 13,609

Force Remove Copilot, Recall and More in Windows 11

remove-ai-watermarks★ 6,470

Remove visible and invisible AI watermarks from images and video, and AI provenance metadata from images, video, and audio. Python library, CLI, and agent skill for SynthID, C2PA, EXIF, IPTC, XMP, and common generative-AI marks. By Victor Kuznetsov.

Free-AppleId-Serve★ 3,685

2026年11月每日更新🚀苹果美区账号 APPLE ID🚀免费Shadowrocket(小火箭)共享账号,Quantumult X,圈x,stash,小火箭节点,🚀 免费订阅地址,🚀 免费节点,🚀 每天更新一次,共享节点,节点质量高可用,完全免费。免费clash订阅地址,免费翻墙,免费科学上网,免费梯子,免费ss/v2ray/trojan/clash节点,翻墙梯子,Openai ChatGPT,代理,proxy代理科学上网,TG代理,电报代理,Telegram代理,免费共享苹果外区ID,clashx Pro客户端下载,Free subscription address,使用教程

conversation-steganography★ 1,271

Use LLMs to hide messages inside normal looking conversations

nostr-vpn★ 1,252

Private mesh VPN for humans and agents. Local keys, signed enrollment, no provider account. Native apps and CLI.

SurrogateShield★ 963

A privacy-first guardrail for LLMs and RAG pipelines. SurrogateShield detects, redacts, and reconstructs sensitive PII data on the fly while defending against adversarial attacks.

pasteguard★ 837

AI gets the context. Not your private data. Local-first privacy proxy for browser chat, AI APIs, and coding agents.

encrypted-note★ 494

Offline, encrypted vault for seed phrases, private keys, passwords and API keys. Also used by agents.

maskit★ 346

数据面具 Data Maskit — 专为大模型打造的本地隐私脱敏网关(请求自动打码,回复流式还原;支持 Cursor / Claude Code / Codex / Pi 等任意可配 Base URL 工具)

Privacy-Data-Protection-Skills★ 344

282+ structured privacy & data protection skills for AI agents. GDPR, CCPA, EU AI Act, HIPAA, LGPD, PIPL, DPDP Act.

mistgate★ 315

Beta. Self-hosted panel for running a VPN for others: family, friends, a community or a small team. Hysteria2, VLESS REALITY and AmneziaWG, one link per person. Add servers from the browser over SSH; client-eye health checks and Telegram alerts; MCP for AI agents. Two static Go binaries, SQLite, no Docker.

wificalling-location-gateway★ 253

Fail-open OpenWrt WLOC location gateway PoC with auditable multi-agent development workflow

veil★ 248

Local privacy proxy for AI coding agents. Keep PII and secrets out of LLMs with deterministic pseudonymization.

VibeGuard★ 235

Uses just 1% memory while protecting 99% of your personal privacy.

gate★ 214

A deterministic privacy boundary between your data and AI.

cloakbot★ 205

A privacy-first AI agent that sanitizes your prompts locally with a local LLM before forwarding to remote LLM APIs.

remove-your-data★ 188

Delete yourself from data brokers for $0. An agent skill: paste the repo URL into Grok Bot, Claude, Codex, Cursor or any coding agent.

pii-masker★ 187

PII Masker is an open-source tool for protecting sensitive data by automatically detecting and masking PII using advanced AI, powered by DeBERTa-v3. It provides high-precision detection, scalable performance, and a simple Python API for seamless integration into workflows, ensuring privacy compliance in various industries.

pebblo★ 150

Pebblo enables developers to safely load data and promote their Gen AI app to deployment

raw-disk-parser★ 110

A tool to interact with Windows drivers to perform a raw disk read and parse out target files without calling standard Windows file APIs

codex-privacy-hud★ 110

OpenAI Privacy hackathon - Paris, 🏆 winning project

rehydra-sdk★ 109

Prevent accidental PII leakage in LLM prompts before they hit the model.

ko-pii★ 107

한국어 PII 검출 + 가명화 Python 라이브러리

CosyRedactGateway★ 107

Lightweight, stateless privacy gateway for LLM APIs — redact secrets before OpenAI/Anthropic upstreams and restore them transparently in SSE and tool calls.

ultrasonic★ 104

A comprehensive steganography framework for embedding and extracting agentic commands in audio and video media using ultrasonic frequencies. This project provides tools for covert communication and command transmission through multimedia channels.

cheap-vpn-airport★ 104

🚁🚀 2026便宜VPN机场推荐 | 性价比机场 | 低价机场!便宜VPN,快连VPN ,小牛VPN,熊猫VPN。低价机场,便宜机场,平价机场,廉价机场,翻墙机场,付费机场,收费机场,高速机场,稳定机场,冷门机场,性价比机场,优质机场,翻墙梯子推荐。解锁AI Agent、Claude、Skills、ChatGPT,YouTube、Netflix、TikTok、bilibili港澳台。翻墙,科学上网,梯子。非免费梯子,非免费VPN,非免费机场!谷歌,油管。适用Clash,V2RAY,小火箭等代理软件。🚀🚁

agentveil★ 88

Your code contains API keys, passwords, and personal data. AgentVeil detects 39 PII & secret types, masks them before AI sees them — then restores on response.

home★ 80

Confidential is the confidential computing stack. We run your AI workloads (inference, training, agents) in hardware-encrypted Trusted Execution Environments (TEEs). You can provide on-prem security and end-to-end privacy guarantees on your off-prem hosted infrastructure. Your data and code stay private. You can independently verify these claims.

datafog-python★ 80

Offline PII firewall for AI agents and LLM apps: fast local detection and redaction, Claude Code hook, LiteLLM guardrail. Zero network calls, one dependency.

home★ 79

Confidential is the confidential computing stack. We run your AI workloads (inference, training, agents) in hardware-encrypted Trusted Execution Environments (TEEs). You can provide on-prem security and end-to-end privacy guarantees on your off-prem hosted infrastructure. Your data and code stay private. You can independently verify these claims.

promptzero★ 70

Zero trace. Full answer. — Transparent Claude API proxy that anonymizes PII and sensitive data before it leaves your environment. From pentesters, to pentesters.

NoPII★ 68

Ready-to-run examples showing NoPII PII protection with OpenAI, Anthropic, LangChain, LlamaIndex, and more

PrivAiTe★ 68

Self-hosted PII redaction proxy for LLM APIs and agent CLIs. Scrubs names, emails and secrets from Claude Code, Codex and any OpenAI-compatible app before requests leave your machine, tool-call arguments included, then restores them in the reply. Local detection, zero telemetry.

onionphone★ 65

VOIP tool for calling over Tor

veil★ 59

使用大模型时,隐藏你的密码、token 等敏感信息。适配市面上各种工具,一键安装即可用。 / Hide passwords, tokens and secrets from the model. Works with everyday tools — one-click install.

PrivacyLens★ 55

A data construction and evaluation framework to quantify privacy norm awareness of language models (LMs) and emerging privacy risk of LM agents. (NeurIPS 2024 D&B)

flare-redact★ 52

Zero-dependency TypeScript secret and PII redaction for any SDK, logs, HTTP, LLM prompts, and scoped MCP tool boundaries.

safe-chat-slack-bot★ 50

SafeChat Slack Bot is an open-source project designed to enhance data security within Slack workspaces.

ai-agent-privacy★ 50

Dataset and evaluation benchmark for Privacy Leakage Evaluation of Autonomous Web Agents

quesma-shipper★ 50

Collects AI coding-agent sessions from developer machines, scrubs secrets, encrypts with age, and uploads to your organisation's storage.

content_analysis_sdk★ 49

This repository contains the SDK that DLP agents may use to become service providers for the Google Chrome Content Analysis Connector.

cover★ 49

Keep private data, internal infrastructure and secrets out of cloud coding agents without breaking your workflow.

iceslab★ 49

Self-hosted proxy panel that runs every protocol on its own upstream core: Xray (VLESS, REALITY, XHTTP), sing-box (TUIC, AnyTLS, ShadowTLS), Hysteria 2, AmneziaWG, NaiveProxy, Shadowsocks 2022, Mieru, MTProto. Cascades, subscriptions, one agent per node.

rosec★ 48

Linux secrets daemon with Secret Service, SSH agent, TOTP, and passkeys. Providers: local vaults, Bitwarden, Proton Pass, KeePassXC, and GNOME Keyring.

impersonate-proxy★ 48

A local MITM proxy that lets you control TLS fingerprints (JA3/JA4), HTTP/2 fingerprints, HTTP header order, and User-Agent — all from a single YAML config file.

membership_inference★ 46

Python package to create adversarial agents for membership inference attacks againts machine learning models

xxd-strip-ai-meta★ 46

Batch-remove AI provenance and image metadata with ExifTool. CLI + Agent Skill, preserving pixel data.

vkpn★ 46

Cross-platform agent to wireguard+vk-turn-proxy

AgentLeak★ 45

AgentLeak (IEEE Access 2026): benchmark for internal-channel privacy leakage in multi-agent LLM systems. Inter-agent messages leak 68.8% vs 27.2% in final outputs; output-only audits miss 41.7%. 1,000 scenarios, 7 channels, 5 LLMs. Tool & MCP server: github.com/yagobski/agentleak

private-connect★ 43

Secure access to private services for developers and AI agents.

z8rdq★ 41

# Shadowrocket: 2021-04-10 03:41:08 [General] bypass-system = true skip-proxy = 192.168.0.0/16, 10.0.0.0/8, 172.16.0.0/12, localhost, *.local, captive.apple.com tun-excluded-routes = 10.0.0.0/8, 100.64.0.0/10, 127.0.0.0/8, 169.254.0.0/16, 172.16.0.0/12, 192.0.0.0/24, 192.0.2.0/24, 192.88.99.0/24, 192.168.0.0/16, 198.18.0.0/15, 198.51.100.0/24, 203.0.113.0/24, 224.0.0.0/4, 255.255.255.255/32 dns-server = system ipv6 = false update-url = https://raw.githubusercontent.com/iSteal-it/script/main/shadowrocket.configuration [Rule] DOMAIN-SUFFIX,baidu.com,DIRECT DOMAIN-SUFFIX,baidubcr.com,DIRECT DOMAIN-SUFFIX,bdstatic.com,DIRECT DOMAIN-SUFFIX,yunjiasu-cdn.net,DIRECT DOMAIN-SUFFIX,taobao.com,DIRECT DOMAIN-SUFFIX,alicdn.com,DIRECT DOMAIN,blzddist1-a.akamaihd.net,DIRECT DOMAIN,cdn.angruo.com,DIRECT DOMAIN,download.jetbrains.com,DIRECT DOMAIN,file-igamecj.akamaized.net,DIRECT DOMAIN,images-cn.ssl-images-amazon.com,DIRECT DOMAIN,officecdn-microsoft-com.akamaized.net,DIRECT DOMAIN,speedtest.macpaw.com,DIRECT DOMAIN-SUFFIX,126.net,DIRECT DOMAIN-SUFFIX,127.net,DIRECT DOMAIN-SUFFIX,163.com,DIRECT DOMAIN-SUFFIX,163yun.com,DIRECT DOMAIN-SUFFIX,21cn.com,DIRECT DOMAIN-SUFFIX,343480.com,DIRECT DOMAIN-SUFFIX,360buyimg.com,DIRECT DOMAIN-SUFFIX,360in.com,DIRECT DOMAIN-SUFFIX,51ym.me,DIRECT DOMAIN-SUFFIX,71.am.com,DIRECT DOMAIN-SUFFIX,8686c.com,DIRECT DOMAIN-SUFFIX,abchina.com,DIRECT DOMAIN-SUFFIX,accuweather.com,DIRECT DOMAIN-SUFFIX,acgvideo.com,DIRECT DOMAIN-SUFFIX,acm.org,DIRECT DOMAIN-SUFFIX,acs.org,DIRECT DOMAIN-SUFFIX,aicoinstorge.com,DIRECT DOMAIN-SUFFIX,aip.org,DIRECT DOMAIN-SUFFIX,air-matters.com,DIRECT DOMAIN-SUFFIX,air-matters.io,DIRECT DOMAIN-SUFFIX,aixifan.com,DIRECT DOMAIN-SUFFIX,akadns.net,DIRECT DOMAIN-SUFFIX,alibaba.com,DIRECT DOMAIN-SUFFIX,alikunlun.com,DIRECT DOMAIN-SUFFIX,alipay.com,DIRECT DOMAIN-SUFFIX,amap.com,DIRECT DOMAIN-SUFFIX,amd.com,DIRECT DOMAIN-SUFFIX,ams.org,DIRECT DOMAIN-SUFFIX,animebytes.tv,DIRECT DOMAIN-SUFFIX,annualreviews.org,DIRECT DOMAIN-SUFFIX,aps.or

shim-cli★ 41

Local traffic visibility and privacy controls for coding agents. See what Claude Code, Codex and Copilot send to the model; mask secrets and personal data where the client allows it.

FirewaLLM★ 40

By calling FirewaLLM, users can ensure the accuracy of the large model while greatly reducing the risk of privacy leakage when interacting with it. We believe that FirewallLLM is a privacy protected chatgpt interaction platform.

pentect★ 38

Local security proxy that keeps secrets and sensitive data out of AI requests. Supports Codex, Claude Code and more.

MEXTRA★ 37

Source code for the ACL'2025 paper titled "Unveiling privacy risks in llm agent memory"

prompt-anonymizer★ 32

A buddy check for your prompts — before they reach an LLM. On-device, reversible anonymization that catches the PII you didn't mean to send.

llm-chat-explorer★ 31

A privacy-focused web app for locally processing and analyzing ChatGPT and Claude chat exports.

llm-differential-privacy-gateway★ 31

Noisegate: a differential privacy gateway that lets an untrusted LLM agent query sensitive data over MCP (Model Context Protocol), with a formal guarantee no individual's record can leak even if the agent is adversarial - enforcement lives in trusted code below the model, validated by a runnable attack gallery.

Passliss★ 29

Passliss is a web application that allows you to generate secure passwords, test the strength of exiting ones and more.

apple★ 28

苹果美区账号 APPLE ID 小火箭免费账号共享 韩国17+19+成人号 ChatGPT 5美金账号 2024年已更新

Redact★ 26

Leverage the benefits of large language models without leaking sensitive information.

vault-engine★ 26

A local-LLM privacy layer for anything you paste into a cloud model — detect, tokenize, restore. Zero deps, reversible, strong on Chinese.

search_privacy_risk★ 25

Code for the paper "Searching Privacy Risks in Multi-Agent Systems via Simulation"

ghost-data-still-exists★ 25

An audit agent that proves whether a person's data is really gone: resolves identity across systems, judges state at the audit cutoff, finds surviving PII and restore paths, and plans the fix. First place out of 4,650 at HackerRank Orchestrate 2026.

pii-tokens★ 22

Claude Code plugin: swaps Brazilian personal data (CPF, CNPJ, email, phone) for reversible tokens before Claude reads it

ruby_llm-top_secret★ 21

Automatically filter sensitive information from RubyLLM conversations using Top Secret.

piighost★ 19

Reversible PII masking for LLM agents: piighost replaces personal data with placeholders before the model sees it, then restores the real values in replies and tool calls. LangChain, Pydantic AI, LlamaIndex and Claude Code integrations, plus an OpenAI/Anthropic-compatible proxy.

dsh-mask★ 18

PII masking middleware for DeepSeek Harness: anonymize names, phones, emails, ID cards, bank cards, keys, and addresses to placeholders before they reach the model, restore them at the display layer, keep the restore table only in memory and a controlled storage domain, never log plaintext, and expose /mask and the mask_test tool

LLMReaper★ 17

Passive LLM Conversation Capture & Sensitive Data Exposure Research

ceil-dlp★ 15

Open-Source DLP for LLMs and Agents that provides Conversational Coherence

pii-compliance-gateway-api★ 15

FastAPI gateway for detecting and sanitizing PII before it reaches downstream systems. Built with LangGraph, Redis, and PostgreSQL.

gettor-bot★ 14

A Signal chatbot to broadcast Tor bridges in countries where Tor relays are blocked/monitored

niobium-skills★ 14

Build apps that compute on data no one can see, not even the server running them. Agent skills that guide your AI coding agents from privacy idea to a working encrypted application.

zkgpt★ 13

Building your GPT-based private knowledge hub using zkSNARK & PLONK

MaskClaw★ 13

MaskClaw项目构建了一套基于端侧多模态大模型的自进化隐私防护架构,通过视觉脱敏与Agentic 规则演进,实现了 Agent 行为偏好与隐私防护策略的动态协同生长,确保了高危隐私数据在云端交互中的物理级安全与个体化适配。

claude-desktop-killswitch★ 13

防止 Claude Desktop 在代理挂掉时从中国大陆 IP 直连 Anthropic —— Windows 防火墙 kill-switch,含零泄漏验证方法

privacy-mask★ 11

Automatically redacts sensitive data in screenshots before sending to AI agents

agentleak★ 11

Open-source privacy testing for AI agents. Python SDK, CLI and MCP tools to detect data leaks across tool calls, memory, inter-agent messages and logs, with redacted reports and CI gates.

secretable★ 10

Telegram as Password Manager

veil-go★ 10

The sensitive data firewall for AI. Detect and mask PII (Emails, Credit Cards, CPFs) locally with zero-latency before sending prompts to LLMs. Thread-safe & Production ready.

chatgpt-shroud★ 10

A Chrome extension enhancing privacy in OpenAI's ChatGPT client. 🛡️ Enables users to easily hide and unhide chat history 🛡️ Operating directly in-browser without accessing or storing chat data. 🛡️ Ideal for maintaining privacy during screen shares 🛡️ Contributions are welcome!

sovereignguard★ 10

Open source AI privacy gateway for GDPR-compliant LLM adoption in EMEA

freedom-stack★ 10

The First Agent Privacy Cloud — Privacy infrastructure for AI agents AND humans. 34 containers, 1 command, zero big tech.

sgkapp★ 10

在线社工库APP

AudAgent★ 10

[PETS'26] AudAgent: Automated Auditing of Privacy Policy Compliance in AI Agents

ai-agent-permissions★ 10

Towards Automating Data Access Permissions in AI Agents

easy-redact★ 10

Easily redact sensitive information from documents.

shade-tree-node★ 10

ShadeNet: anonymous Tor egress for AI agents, zk-proof gated. A local HTTPS proxy and MCP server for the agent; Shade Tree nodes provide the egress. Research preview on Sepolia.

static-ip-verify-guide★ 10

静态住宅IP/原生IP真假鉴别 | 18家实测 | 无赞助·无返佣·不推荐任何商家 | RDAP+ASN+Scamalytics 验证方法

eidolon★ 9

Eidolon is a high-performance reverse proxy designed to secure Large Language Model (LLM) interactions. It intercepts outgoing prompts to providers like OpenAI, Anthropic, or Google, automatically detects and redacts Personally Identifiable Information (PII) using a hybrid engine, and restores the original data in the response.

Septum★ 9

Privacy-first AI middleware with local PII detection, approval gate, and automatic de-anonymization. Query your documents with any LLM — GDPR, KVKK, HIPAA ready.

pii-filter-model-eval★ 9

Comparison case study between Gliner V2 vs OpenAI Privacy filter model for PII Redaction

sphragis★ 9

EU AI Act compliance gateway for Claude Code and OpenAI/Anthropic-compatible LLMs: redacts PII locally and keeps a tamper-evident, hash-chained audit log.

claude-watermark-remover★ 9

Stdlib-first removal of AI provenance marks from text and files (Layer A text scrub, C2PA/EXIF/XMP file cleaners; MIT)

nEigent★ 8

This Project is for HackSecure Hackathon. This is a n8n based Data Leak Prevention AI Agent Project. This scans files from Discord, Slack, GitHub and Drive and checks for Data Leaks.

proxygpt★ 8

Protect your privacy when using LLM chatbots by letting someone else proxy your query for you.

awesome-llm-agent-privacy★ 8

A curated list of papers on privacy, security, and compliance in LLM-based agent systems — attacks, defenses, benchmarks, and regulatory frameworks.

Sec-ConvAgent★ 8

Secure Conversational Agentic-Flow over Agent-User to enhance the Privacy-data Driven Fully Encryption-Decryption Mechanism Conversation between Agent and User

More in Security & AI safety