signoz-explaining-alerts

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFENO_CODECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill is an administrative utility for observability workflows, authored by a known vendor (SigNoz). It operates within the intended scope of interpreting monitoring rules.
  • [COMMAND_EXECUTION]: The skill provides instructions for using vendor-specific MCP tools, including signoz:signoz_get_alert, signoz:signoz_list_alert_rules, and signoz:signoz_get_alert_history. These are used for data retrieval only.
  • [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it processes user-controllable alert metadata (names, descriptions, and annotations). \n
  • Ingestion points: Alert configuration JSON data fetched from signoz:signoz_get_alert. \n
  • Boundary markers: Absent; there are no delimiters defined to isolate external alert data from the agent's instructions. \n
  • Capability inventory: The agent has access to write operations (e.g., signoz_update_alert) and exploratory tools (e.g., signoz-investigating-alerts) as mentioned in the workflow next steps. \n
  • Sanitization: Absent; the skill does not instruct the agent to escape or validate the contents of the fetched alert rule before interpretation.
  • [NO_CODE]: The skill consists entirely of instructional markdown and does not include any accompanying scripts, binaries, or executable logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 08:01 PM