The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides instructions for the agent to process external data, including reconnaissance URL lists and Base64-encoded SAML responses, which serves as a potential surface for indirect prompt injection if the ingested data contains instructions designed to influence the agent.
Ingestion points: Reconnaissance URL lists (recon/$TARGET/urls.txt) and SAML response strings.
Boundary markers: None present.
Capability inventory: Shell command execution using grep, cat, base64, and xmllint as described in SKILL.md.
Sanitization: None present.

hunt-saml