# model_instance_update

A write tool on the Kaggle-MCP MCP server.

Agent View of the PolicyLayer registry record for `model_instance_update`. HTML page: https://policylayer.com/tools/54yyyu-kaggle-mcp/model-instance-update

## Facts

- Tool: `model_instance_update`
- Server: Kaggle-MCP (`54yyyu/kaggle-mcp`) — https://policylayer.com/tools/54yyyu-kaggle-mcp.md
- Homepage: https://github.com/54yyyu/kaggle-mcp
- Risk category: Write (Medium risk)
- Registry record: grade F, identity unverified
- Server rate-limited: no
- Parameters: 0
- Recommended policy verdict: Rate-limited

## Example call (MCP tools/call, JSON-RPC 2.0)

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "model_instance_update",
    "arguments": {}
  }
}
```

## Why model_instance_update is rated Medium

The 'update' verb in the tool name suggests reversible modification of data (a model instance), placing it in the Write category. Given the Kaggle context (competitions, datasets, kernels), this likely modifies user-owned model configuration or state. Severity is medium because unintended modifications could corrupt a user's model setup, but the blast radius is limited to that user's model instance.

From the tool's own definition: "Tool name 'model_instance_update' indicates modification of a model instance. No description provided, requiring inference from name alone."

## Use case

AI agents use model_instance_update to create or update resources in Kaggle-MCP, usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Kaggle-MCP environment.

## Recommended policy (PolicyLayer)

Verdict: **Rate-limited**. Enforced by the PolicyLayer MCP gateway (https://policylayer.com/mcp-gateway) before a call reaches Kaggle-MCP:

```json
{
  "version": "1",
  "default": "deny",
  "tools": {
    "model_instance_update": {
      "limits": [
        {
          "counter": "model_instance_update_rate",
          "window": "minute",
          "max": 30,
          "scope": "grant"
        }
      ]
    }
  }
}
```

## Other tools on Kaggle-MCP (43)

- `config_unset` — Destructive — https://policylayer.com/tools/54yyyu-kaggle-mcp/config-unset.md
- `model_delete` — Destructive — https://policylayer.com/tools/54yyyu-kaggle-mcp/model-delete.md
- `model_instance_delete` — Destructive — https://policylayer.com/tools/54yyyu-kaggle-mcp/model-instance-delete.md
- `model_instance_version_delete` — Destructive — https://policylayer.com/tools/54yyyu-kaggle-mcp/model-instance-version-delete.md
- `authenticate` — Other — https://policylayer.com/tools/54yyyu-kaggle-mcp/authenticate.md
- `competition_details` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/competition-details.md
- `competition_download_files` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/competition-download-files.md
- `competition_leaderboard` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/competition-leaderboard.md
- `competition_list_files` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/competition-list-files.md
- `competition_submissions` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/competition-submissions.md
- `competitions_list` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/competitions-list.md
- `config_view` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/config-view.md
- `dataset_download_files` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/dataset-download-files.md
- `dataset_list_files` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/dataset-list-files.md
- `dataset_metadata` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/dataset-metadata.md
- `dataset_status` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/dataset-status.md
- `datasets_list` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/datasets-list.md
- `kernel_list_files` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/kernel-list-files.md
- `kernel_output` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/kernel-output.md
- `kernel_pull` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/kernel-pull.md
- `kernel_status` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/kernel-status.md
- `kernels_list` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/kernels-list.md
- `model_get` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/model-get.md
- `model_instance_get` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/model-instance-get.md
- `model_instance_list_files` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/model-instance-list-files.md
- `model_instance_version_download` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/model-instance-version-download.md
- `model_instance_version_list_files` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/model-instance-version-list-files.md
- `models_list` — Read — https://policylayer.com/tools/54yyyu-kaggle-mcp/models-list.md
- `competition_submit` — Write — https://policylayer.com/tools/54yyyu-kaggle-mcp/competition-submit.md
- `config_path` — Write — https://policylayer.com/tools/54yyyu-kaggle-mcp/config-path.md
- …and 13 more: https://policylayer.com/tools/54yyyu-kaggle-mcp.md

## For agents

This record is a snapshot. Live verdicts and the full registry:

- Check every server in your MCP config at once: `npx -y policylayer stack`
- Vet a server before you add it: install the mcp-precheck skill — `npx skills add https://policylayer.com` (skill text: https://policylayer.com/skill.md)
- Query the registry over MCP: endpoint `https://api.policylayer.com/mcp` — tools `check_mcp_server`, `check_mcp_stack`, `check_tool`, `search_registry`, `get_change_events`

---

Source: the PolicyLayer MCP registry — one continuously verified record per MCP server. Full record: https://policylayer.com/registry?q=54yyyu-kaggle-mcp · API: https://policylayer.com/registry/api · Policy library: https://policylayer.com/policies/54yyyu-kaggle-mcp