cross-tenant-data-leak-tests

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill includes plaintext database passwords and full connection strings (e.g., PGPASSWORD, POSTGRES_PASSWORD, DATABASE_URL with user:password), so an agent following it would need to handle and reproduce secret values verbatim in commands/configs.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The workflow references remote GitHub Actions that are fetched and executed at CI runtime (actions/checkout@v5 and actions/setup-python@v5), which means external code is pulled and run during the skill's execution.