The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to fetch and process document content from an external Paperless-ngx instance. This introduces a surface for indirect prompt injection where a malicious document could contain instructions intended to influence the agent's behavior.
Ingestion points: The skill reads document contents and search results via the curl ... | jq -r '.content' pattern in SKILL.md and README.md.
Boundary markers: No specific delimiters or safety instructions are provided to the agent to treat the document content as untrusted data.
Capability inventory: The skill utilizes curl for network requests and jq for data processing.
Sanitization: The skill does not implement sanitization or filtering of the fetched document content before presenting it to the agent.

paperless-search