The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection. 1. Ingestion points: Processes external legal texts from EPC, EPO, and PCT sources via RAG search. 2. Boundary markers: No specific delimiters or safety instructions are defined to separate ingested legal text from agent instructions. 3. Capability inventory: The skill has access to Bash, Read, and Write tools, providing significant system access. 4. Sanitization: No evidence of input validation or sanitization for the data retrieved during searches.
[COMMAND_EXECUTION]: The instructions specify using the Bash tool to execute legal search queries. Using a general-purpose shell for search tasks increases the risk of command injection if user-provided search terms are not properly escaped before execution.

epc-search