The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill describes techniques for processing and interacting with external, user-influenced URLs and payloads (webhooks, URL imports, etc.). This introduces a potential surface for indirect prompt injection if the agent fetches and processes content from an attacker-controlled endpoint without proper isolation.
Ingestion points: The skill focuses on processing external data from parameters like url=, uri=, and callback= as defined in the metadata and body.
Boundary markers: No specific delimiters or instructions are provided to the agent to treat fetched external content as untrusted or to ignore embedded instructions within that data.
Capability inventory: The skill utilizes network-capable tools including burp, collaborator, and interactsh to interact with external systems.
Sanitization: The instructions lack guidance on sanitizing or validating the response content from these external requests, focusing instead on identifying successful vulnerabilities.

web-ssrf