Skills
Navigate
skills/jmagar/claude-homelab/homelab-health/Gen Agent Trust Hub

homelab-health

Pass

Audited by Gen Agent Trust Hub on Oct 7, 2026

Risk Level: SAFEDATA_EXFILTRATIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill accesses a sensitive environment file (~/.claude-homelab/.env) to retrieve authentication tokens and API keys.
  • Evidence: ENV_FILE="$HOME/.claude-homelab/.env" and source "$ENV_FILE" in scripts/check-health.sh.
  • [DATA_EXFILTRATION]: Service tokens (e.g., PLEX_TOKEN, RADARR_API_KEY, TAILSCALE_API_KEY) are included in HTTP headers sent to configured endpoints via curl.
  • Evidence: check_service function calls in scripts/check-health.sh include parameters for authentication headers.
  • [COMMAND_EXECUTION]: The skill runs a local bash script to perform its health checks and network operations.
  • Evidence: bash ~/claude-homelab/skills/health/scripts/check-health.sh in SKILL.md.
  • [INDIRECT_PROMPT_INJECTION]: The agent parses external data (service URLs and statuses) from the script output, which presents a surface for potential instruction injection if service metadata is compromised.
  • Evidence: The 'Present the Dashboard' section in SKILL.md instructs the agent to parse the JSON result from check-health.sh.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 7, 2026, 06:27 PM