The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted documents (PDFs, Word documents) which constitutes a surface for indirect prompt injection attacks where content inside the document could attempt to override agent instructions.
Ingestion points: External file content is ingested and parsed via the agentbody_document_parse tool as described in SKILL.md.
Boundary markers: Absent; the instructions do not provide explicit delimiters or warnings to the agent to ignore instructions embedded within the processed documents.
Capability inventory: The skill utilizes sandbase_describe_tool and sandbase_call_tool to interact with the SandBase gateway and execute parsing functions as detailed in SKILL.md and references/sandbase-api-map.md.
Sanitization: Absent; there is no mention of sanitizing, escaping, or validating the extracted text content before it is synthesized into answers for the user.

document-parser