Vishap Security Analyzer for DeepSeek Harness

A DeepSeek Harness tool plugin that brings Vishap — a 6-stage security analysis pipeline — directly into your DSH chat.

🛡️ What it does

When you ask DSH to analyze code, configs, or logs for security risks, this plugin sends the material to a local Vishap backend. Vishap then runs it through its SAST/SCA rules and an LLM to produce a structured report:

• Critical risks (SQLi, XSS, RCE, path traversal, etc.) with CWE IDs and CVSS scores.
• STRIDE threat model.
• Actionable fixes (parameterized queries, input validation, etc.).
• Audit trail with SHA-256 provenance.

🚀 Quick Start

1. Prerequisites

• A running Vishap backend on http://127.0.0.1:5001. (If you don't have it, clone the Vishap repo and run .\.venv\Scripts\python.exe -m uvicorn src.main:app --host 127.0.0.1 --port 5001.)
• DeepSeek Harness (DSH) installed.

2. Install the plugin

dsh plugin --profile desktop add github:dvddv9/my-dsh-plugin

3. Use it

In DSH chat:

Use vishap_analyze to analyze: db.execute(f"SELECT * FROM users WHERE name = '{name}'")

The plugin will return a full security report.

⚙️ Configuration

The plugin expects the Vishap backend to be running on port 5001. If your backend is on a different port, update the URL in src/index.ts and rebuild.

📄 License

MIT