smb-enumeration

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs embedding plaintext credentials into commands and outputs (e.g., echo -n 'PASSWORD', -U 'DOMAIN/USERNAME%PASSWORD', -p 'PASSWORD'), which requires the model to handle and emit secret values verbatim — a direct exfiltration risk.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). This skill contains explicit, deliberate instructions to enumerate SMB shares, download files, and search for credentials (e.g., manspider crawling and mget/put commands and add_credential), which are high-risk data-exfiltration and credential-theft actions despite being framed for authorized pentesting.

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

• Third-party content exposure detected (high risk: 0.85). Step 1/4/6/7 ingest outsider-authored free text from the target host’s SMB shares (e.g., smbclient ... -c 'recurse ON; ...; mget *' and manspider ... which crawls and greps file contents), and that retrieved file text is then placed into the agent’s LLM context via the skill’s evidence/return outputs.