dd-logs
Pass
Audited by Gen Agent Trust Hub on Jun 28, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [SAFE]: The skill exclusively references official Datadog resources, including documentation at docs.datadoghq.com and the datadog-labs GitHub repository. The behavior aligns with the primary purpose of log management.
- [COMMAND_EXECUTION]: The skill executes shell commands via the pup CLI tool and jq to interact with log data (e.g., pup logs search, pup logs archives list). These are standard operations for a Datadog-focused observability skill.
- [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests and processes external log data from the Datadog platform.
- Ingestion points: Log search results retrieved via the pup logs search command.
- Boundary markers: The instructions do not define explicit delimiters to separate ingested log content from the agent's internal reasoning.
- Capability inventory: The skill has the capability to execute shell commands (pup, jq) and process complex JSON data structures.
- Sanitization: While the skill includes a Python example for sanitizing logs at the source, it does not provide sanitization for the results retrieved by the agent's search commands.
Audit Metadata