Plugin Risk Explainer

WHAT IT DOES

Plugin Risk Explainer is a read-only tool that analyzes plugin contracts in modular host systems, generating AI-powered plain-language explanations of what a plugin does, what permissions it has, who installed it, and the risks of those permissions. It reads plugin contract address, host contract address, plugin registration status, plugin permissions, access control lists, function signatures, plugin dependencies, upgrade history, host owner address, privileged function signatures, event logs, transaction history, plugin activation and deactivation status, governance proposals, and plugin version. The tool then presents this data alongside an AI-generated analysis of the plugin's permissions and risks, identifying potential threats such as privileged access, fund exposure, and governance bypass — helping you understand the trustworthiness of a plugin before interacting with it.

HOW TO USE

Connect your wallet using the BGWallet bridge and select your target network. Enter the plugin contract address you want to analyze. Optionally enter the host contract address for context. Click "Analyze plugin risk" to read on-chain contract data. The tool displays the plugin address, host address, registration status, permissions type, and plugin risk level (Low/Medium/High). Detailed information includes installed by, host owner, access control type, function signatures, dependencies, activation status, plugin version, upgrade history, privileged functions, and governance proposals. A risk assessment box provides a summary risk evaluation. An AI explanation appears below the on-chain data, explaining what the plugin does in the host contract, what permissions it has, who installed or approved it, and the risks of those permissions before you interact with the plugin or its functions. The wallet's own signature step still protects you — this explanation is informational only and does not replace reviewing what you approve in your wallet.

TECHNICAL MECHANISM

The tool uses ethers.js v5.7.2 with the provider from the BGWallet bridge to read on-chain contract data. No hardcoded contract addresses are used — you are always in control of which contract you trust. When analyzing plugin risks, the tool reads the plugin contract address from the user input, reads the host contract address from the user input (if provided), checks plugin registration status via the host's plugin registry (e.g., isPluginRegistered()), reads plugin permissions via the host's permission management functions, reads access control lists via the host's access control functions, reads function signatures from the plugin's ABI or bytecode, reads plugin dependencies via the host's dependency management, reads upgrade history by scanning event logs (e.g., PluginAdded, PluginUpgraded), reads host owner address via the host's owner() function, reads privileged function signatures by analyzing function selectors, reads event logs by scanning for plugin-related events, reads transaction history via provider.getTransactionCount(host), reads plugin activation and deactivation status via the host's plugin management functions, reads governance proposals via the governance contract (if available), and reads plugin version via the host's version tracking. The tool fetches event logs by walking backwards in chunks and halving chunk size on failure, ensuring logs are properly read despite varying RPC limitations. The data is formatted and passed to the AI bridge via BGAI.generate() with a structured prompt that includes all the on-chain data read. The AI is instructed to explain what the plugin does in the host contract, what permissions it has, who installed or approved it, and the risks of those permissions — while stating plainly that it cannot detect if a plugin's registered permissions are being used maliciously beyond the on-chain configuration, cannot detect off-chain social engineering attacks or governance manipulation, and cannot guarantee that a plugin's permissions will not be changed in a future upgrade. The AI response is capped at 1,024 tokens to ensure concise, focused plugin risk assessments. All computed values are guarded against negative numbers — if a permission count or function count appears invalid, the tool flags it rather than displaying nonsensical values.

WHAT IT CANNOT SEE

This tool cannot detect if a plugin's registered permissions are being used maliciously beyond the on-chain configuration. It cannot detect off-chain social engineering attacks or governance manipulation. It cannot guarantee that a plugin's permissions will not be changed in a future upgrade. It cannot detect plugins that are registered but have hidden or off-chain functionality. The AI explanation is generated from on-chain data only and cannot detect hidden plugin functionality, off-chain actions, or social engineering attacks. It cannot predict whether a plugin's permissions will be changed or revoked in the future. The tool does not continuously monitor plugin risks; each analysis is a snapshot at the time of the lookup. It cannot detect delayed permission changes, hidden plugins, or off-chain factors. The tool does not store any data persistently; all state is in-memory and resets on reload. The AI explanation is informational and does not constitute financial advice, legal advice, or a guarantee of plugin safety or permission correctness.

PLEASE NOTE

Plugin Risk Explainer works exclusively with EVM-compatible chains. It requires a Web3 wallet (MetaMask or similar) and the BGWallet bridge. Explanations are AI-generated from data the app reads on-chain — always review what you are actually signing in your wallet before approving it. Plugin permissions carry significant risk including privileged access, fund exposure, and governance bypass. Use this tool as a guide only; always cross-check with independent sources and contract audits for critical decisions. This is experimental software; use at your own risk.