Back to skill

Security audit

GKN Phantom

Security checks for vulnerabilities and agentic risk

Overview

This is a powerful penetration-testing skill with real exploit generation and weak runtime guard enforcement in its quick-run path, so it needs manual review before use.

Install only for a controlled, written-authorized penetration-testing workflow. Do not use the quick_combat CLI directly against broad or third-party targets unless you add an external scope guard, lower rate limits, disable exploit/OOB/memory/session persistence features as needed, and review generated PoCs/exploits before running them.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • YARA SignaturesMalware Match, Webshell Match, Cryptominer Match
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (285)

YARA rule 'c2_framework_indicators': Command-and-control framework indicators (Cobalt Strike, Metasploit, Sliver, etc.) [malware]

Critical
Category
YARA Match
Content
#!/usr/bin/env python3
"""CVE Correlator — threat intelligence & vulnerability database correlation (v3.0.0).

Correlates discovered technologies and versions with known CVEs from the
NVD (National Vulnerability Database) and other sources. Provides:
  1. CPE-to-CVE mapping based on technology fingerprinting
  2. Severity-prioritized CVE listing (CVSS v3 scores)
  3. Exploit availability check (Metasploit, ExploitDB, PoC references)
  4. Remediation guidance (patch versions, workarounds)
  5. Risk scoring integration with the main pipeline

This module is deterministic: it takes technology fingerprints and returns
a structured CVE report. It includes a built-in CVE knowledge base for
common web frameworks and middleware.

Usage (CLI):
  python cve_correlator.py --technologies tech.json
  python cve_correlator.py --cpe "cpe:/a:apache:struts:2.5.26"
  python cve_correlator.py --batch targets.json
"""

from __future__ import annotations

import argparse
import json
import re
import sys
Confidence
85% confidence
Finding
YARA rule matched a known malware signature (reverse shell, backdoor, ransomware, C2 framework, or info stealer).

Chaining Abuse

High
Category
Tool Misuse
Content
```bash
# 基础工具链
sudo apt update && sudo apt install -y nmap golang-go

# ProjectDiscovery 全家桶
go install -v github.com/projectdiscovery/nuclei/v3/cmd/nuclei@latest
Confidence
75% confidence
Finding
Tool calls are chained to bypass individual safety checks or escalate capabilities beyond what any single tool call would allow.

Chaining Abuse

High
Category
Tool Misuse
Content
```bash
# 基础工具链
sudo apt update && sudo apt install -y nmap golang-go

# ProjectDiscovery 全家桶
go install -v github.com/projectdiscovery/nuclei/v3/cmd/nuclei@latest
Confidence
75% confidence
Finding
Tool calls are chained to bypass individual safety checks or escalate capabilities beyond what any single tool call would allow.

Intent-Code Divergence

High
Confidence
96% confidence
Finding
The README makes a safety claim that payloads are 'non-destructive' and 'detection-only', but elsewhere explicitly documents exploit generation, OOB validation, active injection, and L4/write-capable stages. This contradiction can mislead operators, reviewers, or downstream agents into underestimating the capability and risk of the skill, increasing the chance of unauthorized or harmful use.

Intent-Code Divergence

High
Confidence
98% confidence
Finding
The contribution guidance requires all payloads to be 'non-destructive (only detection, not exploitation)', yet the skill advertises exploit generation as a core workflow. That inconsistency weakens governance: contributors and users may rely on a false safety boundary while the skill still supports exploit-oriented actions.

Ae1

High
Category
analysis-evasion
Content
`tech_fingerprint.py` does header/body-based detection without httpx;
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
`tech_fingerprint.py` does header/body-based detection without httpx;
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
`tech_fingerprint.py` does header/body-based detection without httpx;
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
`tech_fingerprint.py` does header/body-based detection without httpx;
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
inputSchema: object; // see references/data_schemas.md -> InputSchema
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
inputSchema: object; // see references/data_schemas.md -> InputSchema
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
inputSchema: object; // see references/data_schemas.md -> InputSchema
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
inputSchema: object; // see references/data_schemas.md -> InputSchema
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
inputSchema: object; // see references/data_schemas.md -> InputSchema
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
inputSchema: object; // see references/data_schemas.md -> InputSchema
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
Vulnerability Engine via `scripts/vuln_detector.py`, which generates a
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
Vulnerability Engine via `scripts/vuln_detector.py`, which generates a
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
defined in `references/payload_playbook.md` and `references/advanced_payload_playbook.md` (v3 NEW), and baselined in
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
defined in `references/payload_playbook.md` and `references/advanced_payload_playbook.md` (v3 NEW), and baselined in
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
defined in `references/payload_playbook.md` and `references/advanced_payload_playbook.md` (v3 NEW), and baselined in
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
defined in `references/payload_playbook.md` and `references/advanced_payload_playbook.md` (v3 NEW), and baselined in
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
- `scripts/api_auditor.py` (v3) — API security auditor: GraphQL introspection
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
- `scripts/api_auditor.py` (v3) — API security auditor: GraphQL introspection
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
- `scripts/cloud_security.py` (v3) — Cloud security scanner: S3 bucket
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Ae1

High
Category
analysis-evasion
Content
- `scripts/cloud_security.py` (v3) — Cloud security scanner: S3 bucket
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Static analysis

Detected: suspicious.dynamic_code_execution, suspicious.exposed_resource_identifier, suspicious.exposed_secret_literal (+2 more)

Dynamic code execution detected.

Critical
Code
suspicious.dynamic_code_execution
Location
scripts/waf_evasion.py:347

Plaintext HTTP endpoint targets a CGNAT/Tailscale-range address.

Critical
Code
suspicious.exposed_resource_identifier
Location
scripts/cloud_security.py:140

File appears to expose a hardcoded API secret or token.

Critical
Code
suspicious.exposed_secret_literal
Location
tests/test_quick_combat.py:57

HTTPS certificate verification is disabled.

Warn
Code
suspicious.insecure_tls_verification
Location
scripts/directory_fuzzer.py:629

HTTPS certificate verification is disabled.

Warn
Code
suspicious.insecure_tls_verification
Location
scripts/exploit_generator.py:49

HTTPS certificate verification is disabled.

Warn
Code
suspicious.insecure_tls_verification
Location
scripts/passive_recon.py:213

HTTPS certificate verification is disabled.

Warn
Code
suspicious.insecure_tls_verification
Location
scripts/poc_generator.py:312

HTTPS certificate verification is disabled.

Warn
Code
suspicious.insecure_tls_verification
Location
scripts/quick_combat.py:159

HTTPS certificate verification is disabled.

Warn
Code
suspicious.insecure_tls_verification
Location
scripts/ssl_analyzer.py:234

HTTPS certificate verification is disabled.

Warn
Code
suspicious.insecure_tls_verification
Location
scripts/tech_fingerprint.py:543

Potential obfuscated payload detected.

Warn
Code
suspicious.obfuscated_code
Location
scripts/ssl_analyzer.py:511