implementing-attack-surface-management

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly shows and instructs placing API keys and secrets directly into code and command-line arguments (e.g., shodan.Shodan("YOUR_SHODAN_API_KEY") and --shodan-key/--censys-secret flags), which would require the LLM to handle and output secret values verbatim.

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

• Third-party content exposure detected (high risk: 0.90). This skill explicitly fetches and ingests public, user-generated and third‑party data (e.g., subfinder using certificate-transparency logs/search engines/Wayback/CommonCrawl, httpx probing web pages, Shodan and Censys API queries) as described in SKILL.md and implemented in scripts/agent.py (ServiceFingerprinter.run_httpx, ShodanScanner, CensysScanner, VulnerabilityScanner), and those external results are parsed and used to drive scanning, asset discovery, and scoring—so untrusted web content can materially influence agent actions.