CC Gateway is an AI API identity gateway that normalizes device fingerprints and telemetry to protect user privacy and control data exposure in API proxying.
Source: per README View on GitHub →CC Gateway is gaining attention for addressing privacy concerns in AI API usage by normalizing device identities and environment fingerprints, providing a solution for users who want to control their data exposure without modifying their existing setup. Its unique technical choice is the comprehensive normalization of various identity and environment dimensions, which is not commonly found in similar tools.
Source: Synthesis of README and project traitsRewrites device ID, email, session metadata, and the `user_id` JSON blob in every API request to a single canonical identity, ensuring consistent representation across all interactions.
Source: per READMEReplaces over 40 environment dimensions, including platform, architecture, Node.js version, terminal, package managers, and runtimes, to prevent device-specific information from being exposed.
Source: per READMERewrites system prompts to match the canonical profile, ensuring that no sensitive information is inadvertently included in prompts sent to the API.
Source: per READMERemoves the `x-anthropic-billing-header` to prevent billing information from being transmitted and to enable cross-session prompt cache sharing, reducing costs.
Source: per READMENormalizes process metrics such as physical RAM, heap size, and RSS to canonical values to prevent hardware differences from leaking.
Source: per READMEProvides a single launcher script for clients, eliminating the need for browser OAuth, configuration changes, or additional files.
Source: per READMEManages token refresh internally, ensuring that client machines never contact `platform.claude.com` directly.
Source: per READMEUses existing access tokens on launch and refreshes them silently, ensuring no network calls until necessary.
Source: per READMESupports `HTTPS_PROXY` and `HTTP_PROXY` environment variables for outbound connections, allowing integration with proxy tools like Clash or V2Ray.
Source: per READMEStrips fields like `baseUrl` and `gateway` that could reveal proxy usage in analytics events, enhancing privacy.
Source: per READMEThe architecture of CC Gateway involves a client-side launcher that communicates with a central gateway server. The gateway server handles authentication, token management, and data normalization before forwarding requests to the Anthropic API. It uses TypeScript for development, with dependencies on `https-proxy-agent` and `yaml` for proxying and configuration management. The project utilizes Docker for deployment and includes scripts for setup and client management.
Source: Code tree + dependency filesCenter: project; inner ring: core feature modules; outer ring: key dependencies. Auto-generated from core_features and tech_stack.key_deps.
https-proxy-agentyamlCC Gateway is suitable for users and organizations that require privacy-preserving API proxying, such as those using AI APIs and needing to control the exposure of device and environment information. It is particularly useful in scenarios where multiple devices are used and there is a need to maintain a consistent identity across sessions.
Source: README0.2.0 (no release date provided, no summary of changes)
Source: GitHub ReleasesCC Gateway is a promising project for those seeking to enhance privacy in AI API usage. Its comprehensive approach to data normalization and ease of setup make it a valuable tool for users and organizations concerned about data exposure. However, its alpha status and potential lack of documentation may be barriers for some users.