Sentrint

What it does

Sentrint is an AI tool for reviewing pull requests and finding and fixing bugs. It has a free tier and then charges by usage.

Sentrint checks the GitHub repository behind an application built with an AI coding tool such as Claude, Cursor, Bolt, v0 or Lovable. It looks for hardcoded secrets, database access rules that leave data exposed, vulnerable dependencies and risky code paths, then explains each result in plain English with the file and line where it was found. It reads code only: it does not run the app or test a live site. Repository access is read-only and limited to the single repository chosen when the GitHub App is installed. Each scan runs in its own short-lived Google Cloud container that is destroyed when the scan finishes, and the clone goes with it; only short excerpts around serious findings are kept for a limited period so a fix prompt can be rebuilt. An AI review rechecks the critical and high findings to cut false positives. Each finding comes with a fix prompt written for the tool that built the app, and a rescan shows what changed. Audit logs, a code map and a README grade badge accompany the reports, and a REST API with scoped keys plus an MCP server let coding agents and CI pipelines start scans and read findings. One file can be checked free. Paid scans are sold as one-time packs from $10, with no subscription.

• Security scan of the GitHub repo behind an AI-built app
• Checks for hardcoded secrets and open database rules
• Second AI pass that confirms or discards findings
• Plain-English findings with file and line
• Fix prompts written for the AI tool that wrote the code
• Attack paths mapped to MITRE ATT&CK techniques
• REST API with scoped keys and an MCP server
• Audit logs, code map and README grade badge

What you can do with it

Listing history

1. Nov 3, 2026Checked against sentrint.com
2. Nov 4, 2026Listed