The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the user to add https://rube.app/mcp as an MCP server. This is a third-party service that defines and handles the tool execution logic, acting as an external dependency for the skill's core functionality.
[DATA_EXFILTRATION]: The skill provides tools to access sensitive infrastructure and monitoring data including logs (DATADOG_SEARCH_LOGS), distributed traces (DATADOG_GET_TRACE_BY_ID), and metrics (DATADOG_QUERY_METRICS). Because these tools operate via a third-party bridge (rube.app), this telemetry and potentially its content are exposed to that external service.
[PROMPT_INJECTION]: The skill is vulnerable to Indirect Prompt Injection through ingested log and event data.
Ingestion points: Tools such as DATADOG_SEARCH_LOGS and DATADOG_LIST_EVENTS mentioned in SKILL.md ingest external data into the agent's context.
Boundary markers: None. There are no instructions to the agent to treat external log or event content as untrusted or to ignore potential instructions embedded within them.
Capability inventory: The skill includes several state-changing tools such as DATADOG_CREATE_MONITOR, DATADOG_UPDATE_DASHBOARD, DATADOG_DELETE_DASHBOARD, and DATADOG_CREATE_EVENT which could be misused if the agent is manipulated by injected content.
Sanitization: The skill instructions do not specify any validation or sanitization of retrieved data before processing.

datadog-automation