io.github.Vohlsyr/trustfetch (MCP Server)
The MCP server io.github.Vohlsyr/trustfetch provides prompt-injection scanning and safe webpage fetching for AI agents that read untrusted content. It focuses on helping agents handle untrusted webpages by scanning for prompt-injection risks and fetching content in a safer manner.
🛠️ Key Features
- Prompt-injection scanning
- Safe webpage fetching for AI agents reading untrusted content
🚀 Use Cases
- AI agents retrieving and processing untrusted webpages
- Reducing prompt-injection risk when content is fetched for downstream reasoning
⚡ Developer Benefits
- Integrates injection scanning with the webpage fetching workflow for agents
- Helps ensure agents consume untrusted content more safely
⚠️ Limitations
- No additional capabilities, parameters, or tool details were provided in the available source data.