> ## Documentation Index
> Fetch the complete documentation index at: https://docs.genesiscomputing.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Config Manager Tools

> Centralize and govern environment settings for Genesis Data Agents - knowledge sources, harvester policies, refresh intervals, and auto-harvest behavior - with scoped, auditable configuration.

## **Overview**

### **Tool Name**

```json theme={null}
config_manager
```

### **Purpose**

The **config\_manager** tool group centralizes environment configuration for Genesis Data Agents. It controls user-facing settings like knowledge sources, harvester parameters, refresh intervals, and auto-harvest policies so teams can govern behavior consistently across workspaces and agents.

### **Key Features & Functions**

<Steps>
  <Step title="Centralized settings" icon="sliders">
    Read and update environment settings in a single place
  </Step>

  <Step title="Scoped configuration" icon="gear">
    Apply settings at `GLOBAL`, `WORKSPACE`, or `BOT` scope.
  </Step>

  <Step title="Safe updates" icon="shield-plus">
    Use dry-run and diff previews to validate changes before applying.
  </Step>

  <Step title="Governance alignment" icon="shield-check">
    Enforce enterprise defaults and guardrails.
  </Step>
</Steps>

## **Input Parameters for Each Function**

### **`_config_manager`**

> This tool exposes a single entry point with sub-actions. Exact parameters may vary by deployment; the structure below reflects common usage patterns.

**Parameters**

| **Name** | **Definition** | **Format** |
| - | - | - |
| **action** | Operation to perform. Common values: `LIST`, `GET`, `UPDATE`. | String |
| **scope** | Target scope for configuration. e.g., `GLOBAL`, `WORKSPACE`, `BOT`. | String |
| **key** | Configuration key to read/update (e.g., `knowledge.enabled`, `harvester.refresh_interval_min`). | String |
| **value** | New value to set (for `UPDATE`). JSON-serializable. | Any / JSON |
| **bot\_id** | Target agent when `scope=BOT`. | String |
| **workspace\_id** | Target workspace when `scope=WORKSPACE`. | String |
| **dry\_run** | If `true`, show the planned change without applying. | Boolean |
| **comment** | Change note for audit history. | String |

<Tip>
  Use dry\_run: true with action: `"UPDATE"` to preview diffs before committing changes.
</Tip>

## **Use Cases**

1. Enable/disable auto-harvest globally with standard refresh intervals
2. Toggle knowledge search for a particular data agent
3. Set conservative harvester schedules for large warehouses
4. Apply temporary overrides during maintenance windows

## **Workflow/How It Works**

1. **Read** current values at the intended scope (`LIST`/`GET`).
2. **Propose updates** and run with `dry_run=true` for a preview.
3. **Apply changes** (`UPDATE`) with an audit **comment**.
4. **Validate** downstream behavior (e.g., harvest summaries, tool outputs).

## **Integration Relevance**

* **Harvester Tools:** Align inclusions/exclusions and refresh schedules.
* **Project Manager Tools:** Centralized governance for data agent behaviors used across missions.
* **System Manager Tools:** System-wide controls for stability and maintenance.

## **Configuration Details**

* Prefer **WORKSPACE** scope for team-level adjustments; use **GLOBAL** sparingly.
* Use **dry\_run** to view diffs before applying changes.
* Document rationale in **comment** for compliance and rollbacks.

<Note>
  Scope precedence typically follows `BOT > WORKSPACE > GLOBAL`

  A data agent-level setting overrides workspace/global defaults.
</Note>

## **Limitations or Notes**

* Some settings may be immutable in a managed deployment.
* Invalid keys/scopes or insufficient permissions will be rejected.

## **Output**

* **LIST / GET:** Current settings at the requested scope
* **UPDATE:** Change confirmation with before/after diff (when available)
* **Errors:** Detailed messages for invalid keys, scopes, or permission violations