The Agent Skills Directory

[Data Exposure & Exfiltration] (SAFE): The cryptographic keys provided in the code examples (e.g., sequential digits) are non-functional placeholders intended for documentation purposes and do not constitute actual secret exposure.\n- [Indirect Prompt Injection] (SAFE): The code provides a data ingestion surface that reflects user-provided form data directly into HTML responses, creating a potential Cross-Site Scripting (XSS) vulnerability surface if implemented in production without additional sanitization.\n
Ingestion points: req.parse_form within the handle_form handler in SKILL.md.\n
Boundary markers: None present in the documentation examples.\n
Capability inventory: The code demonstrates network binding via TcpListener and content delivery via res.render.\n
Sanitization: Not implemented in the provided snippets; user input is interpolated directly into HTML strings.

salvo-csrf