offensive-wifi

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). This skill instructs using captured/cracked credentials directly in commands (e.g., -p cracked_pass, --creds, passing passwords on CLI), which requires including secret values verbatim and is an insecure pattern that enables secret exfiltration.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This content is explicitly offensive, providing step‑by‑step instructions for credential theft (evil‑twin/RADIUS, EAP‑TLS cert theft, MSCHAPv2 capture), active network compromise (rogue APs, KARMA/Mana, captive‑portal phishing, deauth/DoS, WPS Pixie Dust, WPA3 downgrades, KRACK/FragAttacks) and post‑compromise pivoting (AD/RADIUS pivots, silentbridge), and therefore carries deliberate malicious intent and very high abuse potential.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly instructs fetching and running public GitHub repositories (e.g., git clone https://github.com/vanhoefm/dragonblood, krackattacks-scripts, fragattacks, silentbridge and similar commands) which are untrusted third‑party code/content that the workflow expects you to read/execute and that could materially change tooling/decisions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs runtime git clone and execution of remote scripts (e.g., https://github.com/vanhoefm/dragonblood, https://github.com/vanhoefm/krackattacks-scripts, https://github.com/vanhoefm/fragattacks, https://github.com/s0lst1c3/silentbridge), meaning it fetches and runs external code during execution and therefore depends on remote content that executes code.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 1.00). Contains explicit sudo commands and instructions to change network interfaces, kill processes, start services and rogue APs (hostapd/eaphammer/silentbridge) which alter system/network configuration and require elevated privileges, so it directs the agent to modify the machine's state.