---
icon: material/monitor-dashboard
title: "Manage"
description: "Manage API keys, environments, licensing, and the metrics your engine publishes"
---

Manage your 10x deployment lifecycle: API keys, environments, authentication,
licensing, and the metrics the engine publishes. Drive it through the
[MCP env and auth tools](../apps/mcp/) or the [REST API](../api/manage.md). API
keys are minted by [MCP sign-in](../apps/mcp/tools/account/signin.md).

<div class="grid cards" markdown>

-   :material-key-outline:{ .lg .middle } **API Key**

    ___

    Generate and rotate the per-environment key the MCP and REST API use.

    [:octicons-arrow-right-24: Learn more](api-keys/)

-   :material-cloud-outline:{ .lg .middle } **Environments**

    ___

    Isolate metrics and data by team, project, or deployment stage.

    [:octicons-arrow-right-24: Learn more](environments/)

-   :material-security:{ .lg .middle } **Identity**

    ___

    Per-environment key for the engine; Auth0 device flow for the MCP.

    [:octicons-arrow-right-24: Learn more](authentication/)

-   :material-license:{ .lg .middle } **License**

    ___

    The signed token the engine verifies offline, and the built-in evaluation license when there is none.

    [:octicons-arrow-right-24: Learn more](license/)

-   :material-heart-pulse:{ .lg .middle } **Status**

    ___

    What status.log10x.com monitors, and how to read it from your own monitoring.

    [:octicons-arrow-right-24: Learn more](status/)

-   :material-text-box-outline:{ .lg .middle } **Logging**

    ___

    Where the engine writes its own logs, and how to change path, appender, and rotation.

    [:octicons-arrow-right-24: Learn more](logging/)

-   :material-robot-outline:{ .lg .middle } **MCP Server**

    ___

    Cost attribution and pattern analysis for AI agents (Claude, Cursor, etc.).

    [:octicons-arrow-right-24: Learn more](../apps/mcp/)

</div>

## Metric output

The engine publishes per-pattern cost and volume metrics to a time-series
backend you own. That store is also the audit trail for what the engine did,
held on your own infrastructure. Pick the output that matches your setup:

| Output | Use case | Configuration |
|--------|----------|---------------|
| Prometheus remote write | your own Prometheus, Grafana Cloud, Mimir, Thanos, VictoriaMetrics | [Remote Write](../run/output/metric/prometheus/remote-write/) |
| Prometheus scrape | pull-based collection | [Scrape](../run/output/metric/prometheus/scrape/) |
| Push Gateway | batch or job workloads | [Push Gateway](../run/output/metric/prometheus/push-gateway/) |
| Datadog | an existing Datadog account | [Datadog](../run/output/metric/datadog/) |
| Elastic | Elasticsearch or OpenSearch | [Elastic](../run/output/metric/elastic/) |
| AWS CloudWatch | AWS-native monitoring | [CloudWatch](../run/output/metric/cloudwatch/) |
| Log4j2 | local files, no time-series database | [Log](../run/output/metric/log/) |

No output is active until you pick one. Full reference:
[Metric Output](../run/output/metric/).