# `list_goal_connections`

Reads your authorized connection summaries for another member's goal. It exposes your own contacts, not the goal owner's complete match list. Reading changes nothing and establishes neither willingness nor outreach.

Annotations: `readOnlyHint: true` · `destructiveHint: false` · `idempotentHint: true` · `openWorldHint: false`

## Inputs

| Input | Required | Type | Description |
| --- | --- | --- | --- |
| `goalId` | yes | string | Exact readable goal UUID owned by another member |
| `pageSize` | no | number | Nonnegative integer; omitted or 0 means 20, capped at 100 |
| `pageToken` | no | string | Returned token; omit or use empty for the first page |

Keep the goal unchanged on continuation. Tokens are private, actor-bound, and expire after one hour. These are live reads. Continue until the token is empty, even after an empty page, before claiming complete coverage.

## Returns

`{ connections: [...], nextPageToken }`. Each connection has the same summary fields as [`list_goal_matches`](/docs/mcp/tools/list_goal_matches). Introduction requests awaiting review come first, followed by other current matches, then sent or declined requests; suggestions and fit priority order each group, with ID breaking ties. Missing or inaccessible goals return `unavailable`; a readable ended goal has no current connections.

Read [`get_goal_match`](/docs/mcp/tools/get_goal_match) for the selected assessment and [`get_person`](/docs/mcp/tools/get_person) with its own-contact `personId` for the person's profile and pending notes. Full [`get_goal`](/docs/mcp/tools/get_goal) includes the first bounded page of these assessments.

## Example

```json list_goal_connections
{ "goalId": "00000000-0000-4000-a000-000000000001", "pageSize": 20 }
```

Result (synthetic data):

```json result:list_goal_connections
{ "connections": [], "nextPageToken": "" }
```
