The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using curl to manage the service registration flow and perform API requests. This includes a loop for polling the registration status and commands for file system management, such as creating directories and writing configuration.
[DATA_EXFILTRATION]: Accesses and writes to a specific configuration file at ~/.allium/credentials to store and retrieve the API key and query IDs. This behavior is confined to the provider's specific configuration requirements. Network communication is directed to the provider's official domain at api.allium.so for data retrieval and authentication.
[PROMPT_INJECTION]: The skill ingests data from external API responses and executes user-provided SQL queries, which creates a surface for potential indirect prompt injection.
Ingestion points: API response bodies and data warehouse results from api.allium.so (documented in SKILL.md and references/custom-sql.md).
Boundary markers: The instructions do not define explicit delimiters or instructions to ignore potential commands embedded within the retrieved data.
Capability inventory: Shell command execution via curl and file system writes.
Sanitization: No specific validation or sanitization of the external data is defined before it is processed by the agent.

allium