zero-trust-security

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFECREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The 'templates/workload-identity.yaml' file contains a hardcoded password ('password') within the SPIRE server's database connection string. While likely intended as a placeholder, it represents a weak default in a security template.
  • [PRIVILEGE_ESCALATION]: The SPIRE agent configuration in 'templates/workload-identity.yaml' specifies 'privileged: true', 'hostPID: true', and 'hostNetwork: true'. These high-privilege settings are necessary for the agent to perform workload attestation on the host node but increase the potential impact of a compromise.
  • [UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: The skill utilizes 'kubectl apply -f' to fetch and execute configuration manifests directly from the official SPIFFE tutorials on GitHub. It also includes the 'scripts/setup-mtls.sh' script which executes local 'openssl' and 'kubectl' commands for automation.
  • [INDIRECT_PROMPT_INJECTION]: The 'scripts/setup-mtls.sh' script accepts Subject Alternative Names (SAN) and service names as command-line arguments that are interpolated into shell commands. This creates an attack surface for indirect injection if an agent sources these values from untrusted external data.
  • Ingestion points: Command-line arguments in 'scripts/setup-mtls.sh'.
  • Boundary markers: Absent.
  • Capability inventory: Certificate generation via 'openssl' and Kubernetes secret management.
  • Sanitization: No input validation is performed on the provided arguments before use in shell commands.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 05:06 PM
Security Audit — agent-trust-hub — zero-trust-security