🧪 Secrets.Sandbox

This repository is the sandbox target for the MCP Secrets Detector project — an AI-powered tool that scans GitHub Pull Requests for hardcoded secrets and posts inline review comments automatically.

---

🔗 Connection with MCP.Secrets.Detector

Detail	Value
Detector Repo	MCP.Secrets.Detector
Sandbox Repo	davidCrejenovschi/Secrets.Sandbox (this repo)
Branch with secrets	add-dummy-secrets
File with dummy secrets	file.py
Target PR	Opened from add-dummy-secrets into main

The detector tool connects to a Pull Request opened from the add-dummy-secrets branch, scans the changed files, and posts line-level review comments flagging each secret.

---

⚠️ All Secrets Are Fake

Every credential, token, key, password, and connection string in this repository is entirely made up. They are designed to look realistic so the MCP Secrets Detector can be demonstrated and tested, but none of them grant access to any real system:

• The AWS keys are synthetically generated — they match the AKIA prefix format but are not registered with AWS
• The OpenAI API key follows the sk-proj- format but is not associated with any account
• The database URL points to a non-existent host (db-prod-instance.c9xyz.eu-central-1.rds.amazonaws.com) with fake credentials
• The username/password pair (root / SuperSecretAdminPassword123!) is fictional

---

📂 Branch Structure

• main branch — clean, contains only this README. No secrets, no file.py.
• add-dummy-secrets branch — contains the dummy file.py with fake credentials used for detector testing. A Pull Request is opened from this branch into main so the MCP Secrets Detector can scan the diff and post inline review comments.