The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion and processing of data from external API endpoints, creating a potential vector for indirect prompt injection.
Ingestion points: Data is fetched from external URLs via the BackendApiService class in SKILL.md (specifically within getUser and getMessages methods).
Boundary markers: The skill utilizes structured JSON deserialization with kotlinx.serialization, which provides technical delimiters, but lacks explicit sanitization for natural language instructions that might be embedded in the API responses.
Capability inventory: The skill utilizes the Ktor HttpClient in SKILL.md for full network operations, enabling the agent to read from and write to remote services.
Sanitization: While the data is mapped to typed DTOs (User, SavedMessage), there is no mechanism provided to filter or escape the raw text content for potential prompt injection before it is processed by the agent.

ktor-client