The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from external API responses, including job logs and stream events, which are considered untrusted content sources. This establishes an indirect prompt injection surface area.
Ingestion points: Data is ingested through API calls such as GET /v2/job/{id}/logs and GET /v2/stream/{stream}/events as described in health-check.md and streams.md.
Boundary markers: The instructions do not provide explicit delimiters or instructions for the agent to ignore potentially malicious content within these data streams.
Capability inventory: The skill has the capability to execute shell commands (curl) for network operations.
Sanitization: There are no explicit sanitization routines defined for the processing of event or log data before it is presented to the agent.
[COMMAND_EXECUTION]: The skill provides the agent with shell commands (curl) to interact with Lytics API endpoints. These commands are strictly scoped to the Lytics service and are necessary for the skill's monitoring functionality.
[DATA_EXPOSURE]: The instructions describe accessing a local configuration file at ~/.lytics/accounts.toml to manage API tokens for multi-account support. This is a standard credential management pattern for Lytics tooling and is used to facilitate legitimate vendor functionality.

lytics-data-health