# create_completion

Create OpenAI completion

Agent View of the PolicyLayer registry record for `create_completion`. HTML page: https://policylayer.com/tools/io-github-portel-dev-ncp/create-completion

## Facts

- Tool: `create_completion`
- Server: Ncp (`@portel/ncp`) — https://policylayer.com/tools/io-github-portel-dev-ncp.md
- Install: `npx -y @portel/ncp`
- Homepage: https://github.com/portel-dev/ncp
- Risk category: Execute (High 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": "create_completion",
    "arguments": {}
  }
}
```

## Why create_completion is rated High

This tool invokes an external service (OpenAI API) to generate a completion. It is not a simple read (it sends data and triggers computation), nor does it write/delete local data. It falls under Execute because it triggers an external operation whose effects depend on the arguments passed (prompt, model, parameters). Misuse could lead to unintended content generation, prompt injection, or unexpected API costs.

From the tool's own definition: "'Create OpenAI completion' — triggers an external API call to OpenAI, executing a language model inference operation"

## Use case

AI agents invoke create_completion to trigger actions in Ncp. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call: builds kicked off, notifications sent, workflows started.

## Recommended policy (PolicyLayer)

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

```json
{
  "version": "1",
  "default": "deny",
  "tools": {
    "create_completion": {
      "limits": [
        {
          "counter": "create_completion_rate",
          "window": "minute",
          "max": 10,
          "scope": "grant"
        }
      ]
    }
  }
}
```

## Other tools on Ncp (274)

- `backup` — Destructive — https://policylayer.com/tools/io-github-portel-dev-ncp/backup.md
- `bulk_delete` — Destructive — https://policylayer.com/tools/io-github-portel-dev-ncp/bulk-delete.md
- `delete` — Destructive — https://policylayer.com/tools/io-github-portel-dev-ncp/delete.md
- `delete_event` — Destructive — https://policylayer.com/tools/io-github-portel-dev-ncp/delete-event.md
- `delete_key` — Destructive — https://policylayer.com/tools/io-github-portel-dev-ncp/delete-key.md
- `delete_memory` — Destructive — https://policylayer.com/tools/io-github-portel-dev-ncp/delete-memory.md
- `purge_cache` — Destructive — https://policylayer.com/tools/io-github-portel-dev-ncp/purge-cache.md
- `remove` — Destructive — https://policylayer.com/tools/io-github-portel-dev-ncp/remove.md
- `remove_container` — Destructive — https://policylayer.com/tools/io-github-portel-dev-ncp/remove-container.md
- `write_query` — Destructive — https://policylayer.com/tools/io-github-portel-dev-ncp/write-query.md
- `apply` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/apply.md
- `apply_manifest` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/apply-manifest.md
- `build_image` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/build-image.md
- `build_job` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/build-job.md
- `bulk_operations` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/bulk-operations.md
- `click_element` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/click-element.md
- `clone` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/clone.md
- `connect` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/connect.md
- `consume` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/consume.md
- `create_context` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/create-context.md
- `create_image` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/create-image.md
- `create_prediction` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/create-prediction.md
- `create_video` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/create-video.md
- `deploy` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/deploy.md
- `deploy_project` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/deploy-project.md
- `deploy_server` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/deploy-server.md
- `deploy_site` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/deploy-site.md
- `exec` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/exec.md
- `exec_async` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/exec-async.md
- `execute` — Execute — https://policylayer.com/tools/io-github-portel-dev-ncp/execute.md
- …and 244 more: https://policylayer.com/tools/io-github-portel-dev-ncp.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=io-github-portel-dev-ncp · API: https://policylayer.com/registry/api · Policy library: https://policylayer.com/policies/io-github-portel-dev-ncp