observability-dashboard

Warn

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill requires the agent to generate and install multiple executable hook scripts into the .claude/hooks/ directory to monitor various lifecycle events.
  • [DATA_EXFILTRATION]: The pipeline captures and transmits session data, tool history, and agent transcripts via HTTP POST to a central server, creating a pathway for sensitive execution data to leave the immediate agent context.
  • [PROMPT_INJECTION]: The dashboard processes and displays untrusted data from multiple agents (transcripts and tool results) without specified sanitization, creating a surface for indirect prompt injection.
  • Ingestion points: Event JSON payloads, tool call results, and session transcripts from the workflow.
  • Boundary markers: The instructions do not define delimiters or isolation protocols for data displayed in the dashboard UI.
  • Capability inventory: File system writes for hook installation, subprocess management for the monitoring server, and network communication via WebSockets and HTTP.
  • Sanitization: No sanitization or verification steps are included for the captured data before transmission or display.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 17, 2026, 10:52 PM
Security Audit — agent-trust-hub — observability-dashboard