Skills
skills/datadog-labs/agent-skills/agent-observability-experiment-py-bootstrap/Gen Agent Trust Hub

agent-observability-experiment-py-bootstrap

Pass

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill performs local environment introspection by executing shell commands like grep and bash to discover LLM integration points within the user's codebase. It also uses python for syntax validation of the generated experiment files to ensure they are runnable.
  • [DATA_EXFILTRATION]: Upon each invocation, the skill sends a telemetry beacon to Datadog's infrastructure via either an MCP tool call or the pup CLI. This beacon includes a unique invocation ID and intent tags used for usage attribution and connectivity validation, as described in the documentation.
  • [EXTERNAL_DOWNLOADS]: The skill directs users to install and use official Datadog libraries and tools, specifically the ddtrace SDK and the pup CLI. These resources are provided by a well-known and reputable technology organization for the skill's intended purpose.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 19, 2026, 08:57 AM
Security Audit — agent-trust-hub — agent-observability-experiment-py-bootstrap