Troubleshoot Policy

Start with the failed task, not a service restart. Policy tracks draft work, governed revisions, and protection progress separately. A healthy service does not prove that a policy protects its targets.

Choose Your Next Task

Use this table to choose the troubleshooting guide that matches the symptom:

Before Changing Anything

Complete the following steps:

1. Retain the policy identifier, failed action, timestamp, and returned message. Include the time zone and any support reference.
2. Read the policy’s current state through your permitted interface. Record draft status separately from governance and protection states.
3. Confirm who can perform the next task. Authors, reviewers, approvers, auditors, and system operators have different authority.
4. Open the matching playbook above. Each playbook defines its evidence, stop conditions, and success checks.

Do not share credentials or invent another user’s audit identity. Owner or editor access does not grant every command capability. Refer to roles and permissions.

When You Cannot Inspect the Evidence

An ordinary author might lack access to restricted diagnostics or operations. Retain the visible error and ask an authorized operator to investigate. A support reference does not grant diagnostic access. Policy does not provide universal self-service recovery for audit-only evidence. Use the glossary to distinguish authentication tokens, version validators, and delivery evidence.

Use the diagnostic handoff when the next action requires authority you do not have. Keep the recovery directory for links to earlier recovery topics.