Status: [LIVE] Endpoints found: https://api.anthropic.com/v1/messages Source: https://github.com/DungNguyen0209/aibodyguard

AIBodyguard

Credential leak prevention for AI coding agents.

CI Release License Go version

--- ![AIBodyguard in action](docs/assets/demo.png) --- ## Architecture ![AIBodyguard Architecture](docs/assets/architect.png) --- AIBodyguard sits between your AI coding agent and the LLM API. It scans your project for credential files at startup, then intercepts every outbound HTTPS request and redacts any discovered secrets — API keys, database passwords, JDBC URLs — before they leave your machine. No agent configuration needed. Just prefix your command. ```bash aibodyguard claude aibodyguard opencode ``` --- ## Minimum System Requirements | | Minimum | Recommended | |---|---|---| | **OS** | macOS 13 (arm64), Linux (amd64/arm64), Windows 10 (amd64) | macOS 14+, Ubuntu 22.04+ | | **CPU** | Any 64-bit | Apple M-series or modern x86-64 | | **RAM** | 512 MB free | 1 GB+ free | | **Disk** | 400 MB free | 500 MB+ free | | **Network** | Internet access on first run | — | > [!NOTE] > On first run, AIBodyguard downloads the ML model (~67MB) and ONNX Runtime library (~220MB) to `~/.cache/aibodyguard/`. Subsequent runs are fully offline. Total disk usage after setup: ~290MB. --- ## Installation ### macOS ```bash brew install DungNguyen0209/tap/aibodyguard ``` ###