Operate from the Command Center

The Command Center is the first place to understand what needs attention in the active organization. It combines pending approvals, recent activity, policy outcomes, and links to deeper investigation.

What to check first

Read the page in this order:

1. Pending approvals: actions waiting for a person or approaching expiry.
2. Runtime activity: changes in request volume, review demand, or blocked actions.
3. Recent audit events: the latest decisions, configuration changes, and reason details.
4. Inventory and control links: the agent, MCP server, policy, or setting that needs follow-up.

Summary counts tell you where to look. Open the underlying record before making a decision or reporting an incident.

Find a destination quickly

Press Command/Control + K from the authenticated product to search product areas, registered agents, governed tools, and MCP servers. The search helps you navigate; it does not approve requests, change policy, or modify inventory.

Search results belong to the active organization. Confirm the current workspace before opening or acting on a record with a similar name.

Review an approval

Open a pending item and verify:

• who or what requested the action;
• the exact operation, target, and environment;
• the stated risk and data sensitivity;
• which policy or reason required review;
• the request's age and expiry; and
• the expected recovery path if the action fails.

Approve only the action shown. Do not treat approval of one request as permission for future actions with different context.

Understand activity trends

Use runtime volume and policy outcomes to spot:

• a new deployment sending more actions than expected;
• a sudden increase in blocked or reviewed requests;
• repeated retries from one agent;
• activity outside the expected operating window; or
• a quiet integration that should be sending decisions.

Charts reflect traffic that reached Sernixa. Compare unusual gaps or spikes with application and destination-system telemetry.

Investigate an unexpected decision

1. Open the event and read the structured reason.
2. Confirm the organization, caller identity, agent, operation, and environment.
3. Check the matching registry record and policy version.
4. Review nearby decisions for the same caller or target.
5. Use the Trust Graph to review explicitly authored access paths.
6. Use Flight Recorder when runtime evidence is connected.
7. Confirm the final outcome in the system that owned execution.

Daily operating routine

• Clear or expire pending approvals deliberately.
• Investigate changes in block, review, and error rates.
• Check ownership on newly registered agents and MCP servers.
• Review credential, membership, and policy changes.
• Record remediation in the system your team uses for incident follow-up.

Next step

Set the expected decision posture in Policies and controls, improve ownership in the Agent inventory, or review access paths in the Trust Graph.