dd-audit-ai-activity
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill defines several shell commands using 'pup' (Datadog's CLI) and 'jq' to search and process audit logs. These commands are used as intended for auditing system activity and do not exhibit malicious behavior.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and summarize external data from Datadog audit logs, which presents a potential surface for indirect prompt injection.
- Ingestion points: Audit log data is retrieved from the Datadog API via the 'pup audit-logs search' command and processed as JSON.
- Boundary markers: The provided command templates do not include specific delimiters or instructions to the agent to ignore potentially malicious content within the log data during the summarization phase.
- Capability inventory: The skill uses shell execution capabilities ('pup', 'jq') to interact with the environment and Datadog APIs.
- Sanitization: Use of 'jq' ensures the data is processed as structured JSON, though it does not filter the semantic content of individual log fields. This exposure is consistent with the skill's primary purpose of log analysis and is considered a low-risk surface.
Audit Metadata