> Fetch clean Markdown by appending `.md` to any page URL under https://signalwire.com/docs or requesting it with the HTTP header `Accept: text/markdown`. The root index at https://signalwire.com/docs/llms.txt lists the available documentation indexes.

# InfoGathererSkill

> Gather answers to a configurable list of questions, one at a time.

[add-skill]: /docs/server-sdks/reference/typescript/agents/agent-base/add-skill

Collect answers to a configured list of questions. The skill registers a
`start_questions` tool and a `submit_answer` tool; the agent advances through
the question list and reports completion when every question is answered.

Fails closed during `setup()` when `questions` is missing or invalid.

**Class:** `InfoGathererSkill`

**Tools:** `start_questions`, `submit_answer` (names are prefixed when
`prefix` is set)

**Env vars:** None

**Multi-instance:** yes

**`questions`** `QuestionDefinition[]` — required

Ordered list of questions to ask. Each question has:

* `key_name` (string, required) — key used to store the caller's answer.
* `question_text` (string, required) — the question to ask.
* `confirm` (boolean) — when `true`, the agent asks the caller to confirm
  the answer before moving on.
* `prompt_add` (string) — optional extra prompt text appended while this
  question is active.

---

**`prefix`** `string`

Optional prefix for tool names and the global-data namespace. When set,
tools are named `<prefix>_start_questions` / `<prefix>_submit_answer` and
state is stored under `skill:<prefix>`. Required when loading multiple
`InfoGathererSkill` instances on the same agent.

---

**`completion_message`** `string`

Message returned after every question has been answered. Defaults to a
generic completion message.

---

## Example

```typescript {6-14}
import { AgentBase, InfoGathererSkill } from '@signalwire/sdk';

const agent = new AgentBase({ name: 'intake', route: '/intake' });
agent.setPromptText('Collect contact information for follow-up.');

await agent.addSkill(new InfoGathererSkill({
  questions: [
    { key_name: 'full_name', question_text: 'What is your full name?' },
    { key_name: 'email', question_text: 'What is your email address?', confirm: true },
    { key_name: 'phone', question_text: 'What is your phone number?' },
  ],
  completion_message: 'Thanks — we will be in touch shortly.',
}));

agent.run();
```

> Gather answers to a configurable list of questions, one at a time.