bb-local-toolkit

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill includes PoC templates and command-line examples that place secrets/tokens/cookies directly into requests (e.g., Authorization/Cookie headers and curl/workflow exfil patterns) and explicitly asks for "exact HTTP request" reproduction, which would require the LLM to emit secret values verbatim.

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.90). These URLs include attacker-controlled domains, open-redirect/obfuscation examples and cloud storage / Firebase patterns commonly used to host or redirect to malicious payloads, so they should be treated as suspicious download/distribution sources.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This document contains explicit, actionable techniques for secret exfiltration, remote code execution, persistent backdoors (runner/docker escalation), and supply-chain poisoning—high-risk abuse capabilities if used maliciously.

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

• Attempt to modify system services in skill instructions detected (high risk: 0.80). The skill contains explicit post‑exploitation instructions (e.g., privileged docker mount to host, "add SSH keys, modify sudoers, install persistent backdoors", dumping process memory) that instruct an agent to modify host system state and persist access, which would require sudo/privileged operations.