The Agent Skills Directory

[PROMPT_INJECTION]: The skill provides an attack surface for indirect prompt injection by enabling the agent to ingest and act upon content retrieved from an external database via MCP tools.
Ingestion points: Data returned from spacetimedb.sql, spacetimedb.get_schema, and spacetimedb.call tools enter the agent's context during operation.
Boundary markers: The instructions do not specify any delimiters or safety warnings to help the agent distinguish between its core instructions and data content fetched from the database.
Capability inventory: The agent is granted the ability to inspect database schemas, execute SQL queries, and trigger state changes through database reducers.
Sanitization: The skill lacks explicit instructions for sanitizing, escaping, or validating external database content before it is processed by the agent's language model.

mcp