bug-bounty-methodology

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs agents to extract and document secrets (e.g., "extract... secrets", "help capture authentication details", and store "Testing Accounts: Credentials") into project files, which requires the LLM to handle and potentially output secret values verbatim.

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

• Third-party content exposure detected (high risk: 0.72). The required runtime workflow spawns parallel agents that execute recon steps like subdomain enumeration and technology fingerprinting (e.g., crt.sh/subfinder/amass/whatweb) and then synthesize results by reading the agents’ output files; those recon tools can fetch public web content and ingest scraped page/text into the agents’ LLM context, creating an outsider free-text injection surface.