troubleshoot-apache-httpd

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill implements structured troubleshooting procedures for Apache HTTPD by querying signals through a Multi-Context Protocol (MCP) interface. It focuses on observability and standard triage workflows without introducing dangerous command execution or unauthorized data access.
  • [PROMPT_INJECTION]: Indirect Prompt Injection Surface: The skill instructs the agent to ingest external data (metrics, scoreboard states, and log-derived signals) through MCP tools, which is a standard requirement for monitoring tasks but establishes a theoretical injection surface.
  • Ingestion points: Data is retrieved via query_metrics, find_anomalous_metrics, and list_raised_alerts tools as described in SKILL.md and the rules/ subdirectory.
  • Boundary markers: The instructions do not specify delimiters or directions to ignore instructions potentially embedded in the monitored data or log snippets.
  • Capability inventory: The agent has the capability to query diverse metric contexts and recommend service remediations based on the findings.
  • Sanitization: No explicit sanitization or validation of the content returned by the Netdata MCP tools is required by the skill instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 11:24 AM
Security Audit — agent-trust-hub — troubleshoot-apache-httpd