signoz-investigating-alerts
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill implements a diagnostic workflow using SigNoz platform tools to investigate alert fires without modifying system state. All external references (signoz-mcp-setup) and tool naming conventions (signoz_*) align with the verified vendor identity (signoz).- [INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection attack surface as it ingests untrusted data from application logs and trace status messages during the Tier 3 analysis phase. • Ingestion points: Untrusted log bodies and trace messages retrieved via signoz_search_logs and signoz_search_traces. • Boundary markers: Not explicitly defined in the prompt interpolation, although the workflow instructs the agent to group and summarize the data. • Capability inventory: Limited to read-only SigNoz MCP tools (metrics, logs, traces). • Sanitization: No explicit sanitization or escaping of telemetry content is described before data is presented to the agent context.
Audit Metadata