The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a bash launcher script to locate and execute a local binary file on the host system to provide its MCP tool functionality.
Evidence: exec calls in scripts/select-binary.sh target either an environment-specified binary or platform-suffixed binaries in predefined asset directories.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external databases and uploaded files to perform analysis, which could contain adversarial instructions targeting the agent.
Ingestion points: Data enters the agent's context through the query and file_id parameters in data_agent_create_session, and the message parameter in data_agent_send (SKILL.md).
Boundary markers: Absent. The instructions do not provide delimiters or warnings for the agent to ignore instructions embedded in the external data.
Capability inventory: The skill possesses significant capabilities including database querying, file generation (Excel/PDF), and session state management via data_agent_* tools.
Sanitization: Absent. There is no evidence of sanitization or filtering of the content retrieved from external data sources before it is processed by the agent.

alibabacloud-data-agent-mcp-skill