# runner_auth_browser_start

Start the interactive plan-OAuth browser/device-auth login flow for Claude Code or Codex on the local or a remote Yaver machine. Uses the user's Claude Pro/Max or ChatGPT Plus/Pro account on their own device; no API key is requested.

Agent View of the PolicyLayer registry record for `runner_auth_browser_start`. HTML page: https://policylayer.com/tools/io-github-kivanccakmak-yaver/runner-auth-browser-start

## Facts

- Tool: `runner_auth_browser_start`
- Server: Yaver (`yaver-cli`) — https://policylayer.com/tools/io-github-kivanccakmak-yaver.md
- Install: `npx -y yaver-cli`
- Homepage: https://github.com/kivanccakmak/yaver.io
- Risk category: Execute (High risk)
- Registry record: grade F, identity unverified
- Server rate-limited: no
- Parameters: 2 (1 required)
- Recommended policy verdict: Rate-limited

## Parameters

| Parameter | Type | Required | Description |
| --- | --- | --- | --- |
| `runner` | string | yes |  |
| `device_id` | string | no | Optional remote device ID |

Parameters from the server's own tool schema.

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

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "runner_auth_browser_start",
    "arguments": {
      "runner": "<runner>"
    }
  }
}
```

## Why runner_auth_browser_start is rated High

This tool executes an authentication flow that interacts with external services (Claude/ChatGPT OAuth providers) and modifies local/remote machine state (login credentials, session tokens). While ostensibly for legitimate user auth, an AI agent given control could initiate unwanted authentication flows, potentially on remote machines, or use this as a vector to establish persistent access.

From the tool's own definition: "'Start the interactive plan-OAuth browser/device-auth login flow' — initiates an external OAuth authentication process with side effects dependent on user interaction and network state."

## Use case

AI agents invoke runner_auth_browser_start to trigger actions in Yaver. 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 Yaver:

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

## Other tools on Yaver (644)

- `acl_remove_peer` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/acl-remove-peer.md
- `cancel_schedule` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/cancel-schedule.md
- `cloud_destroy` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/cloud-destroy.md
- `companion_down` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/companion-down.md
- `data_delete` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/data-delete.md
- `db_push` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/db-push.md
- `db_reset` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/db-reset.md
- `db_restore` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/db-restore.md
- `dns_flush` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/dns-flush.md
- `dns_remove` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/dns-remove.md
- `docker_prune` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/docker-prune.md
- `docker_rm` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/docker-rm.md
- `docker_rmi` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/docker-rmi.md
- `feedback_delete` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/feedback-delete.md
- `forgot_password` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/forgot-password.md
- `git_member_remove` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/git-member-remove.md
- `git_stash` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/git-stash.md
- `guest_delete` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/guest-delete.md
- `guest_revoke` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/guest-revoke.md
- `jobs_cancel` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/jobs-cancel.md
- `machine_onboarding_remove` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/machine-onboarding-remove.md
- `machine_remove` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/machine-remove.md
- `migrate_rollback` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/migrate-rollback.md
- `models_remove` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/models-remove.md
- `monitor_remove` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/monitor-remove.md
- `phone_project_delete` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/phone-project-delete.md
- `proxy_remove` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/proxy-remove.md
- `relay_clear_password` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/relay-clear-password.md
- `release_rollback` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/release-rollback.md
- `remote_destroy` — Destructive — https://policylayer.com/tools/io-github-kivanccakmak-yaver/remote-destroy.md
- …and 614 more: https://policylayer.com/tools/io-github-kivanccakmak-yaver.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-kivanccakmak-yaver · API: https://policylayer.com/registry/api · Policy library: https://policylayer.com/policies/io-github-kivanccakmak-yaver