> **Building with AI coding agents?** Install the authstack plugin with one command. This equips your agent with accurate Scalekit implementation patterns.
>
> **Recommended** (the `-y` flags skip prompts, so the command runs without an interactive terminal):
> ```bash
> npx -y @scalekit-inc/cli setup -y
> ```
>
> Global:
> ```bash
> npm install -g @scalekit-inc/cli
> scalekit setup -y
> ```
>
> Supports Claude Code, Cursor, GitHub Copilot, Codex + skills for other Agent Skills-compatible agents in `~/.agents/skills`.
> Skills: integrate-agentkit, implement-saaskit, add-mcp-oauth, implement-sso, implement-scim.
> [Full setup guide](https://docs.scalekit.com/dev-kit/build-with-ai/)

---

# Connect AI agents to LLMWhisperer

The LLMWhisperer connector lets your AI agent act in each user's LLMWhisperer account. Each user connects their own LLMWhisperer API key once, and Scalekit sends it with every call, so your agent never handles credentials. It comes with 13 tools.

**Authentication:** API Key
**Categories:** AI, Files & Documents
**Tools:** 13: 9 read, 3 write, 1 destructive
**Users sign in with:** API key
**Built by:** Scalekit connector
**Try it:** [Playground in the Scalekit dashboard](https://app.scalekit.com/ws/signup?sk_intent=playground&provider=LLMWHISPERER)

## Setup

1. ### Install the SDK

   **Node.js**

   ```bash
   npm install @scalekit-sdk/node dotenv
   ```

   **Python**

   ```bash
   pip install scalekit-sdk-python python-dotenv
   ```

2. ### Set your credentials

   Add your Scalekit credentials to your `.env` file. Find values in **[app.scalekit.com](https://app.scalekit.com)** > **Developers** > **API Credentials**.

   ```sh title=".env"
   SCALEKIT_ENVIRONMENT_URL=<your-environment-url>
   SCALEKIT_CLIENT_ID=<your-client-id>
   SCALEKIT_CLIENT_SECRET=<your-client-secret>
   ```

3. ### Create the LLMWhisperer connection

   In **AgentKit > Connections**, create a LLMWhisperer connection. The name you give it is the `connection_name` your code passes. See [Configure connections](/agentkit/connections/).

4. ### Authorize a user and make your first call

   **Node.js** (`quickstart.mts`)

   ```typescript
   import { ScalekitClient } from '@scalekit-sdk/node'
   import 'dotenv/config'
   import { createInterface } from 'node:readline/promises'

   const scalekit = new ScalekitClient(
     process.env.SCALEKIT_ENVIRONMENT_URL,
     process.env.SCALEKIT_CLIENT_ID,
     process.env.SCALEKIT_CLIENT_SECRET,
   )
   const actions = scalekit.actions

   const connector = 'llmwhisperer'
   const identifier = 'user_123'

   // Generate an authorization link for the user
   const { link } = await actions.getAuthorizationLink({ connectionName: connector, identifier })
   console.log('Authorize LLMWhisperer:', link)
   const rl = createInterface({ input: process.stdin, output: process.stdout })
   await rl.question('Press Enter after authorizing...')
   rl.close()

   // Make your first call
   const result = await actions.executeTool({
     connector,
     identifier,
     toolName: 'llmwhisperer_usage_info_get',
     toolInput: {},
   })
   console.log(result)
   ```

   ```bash
   npx tsx quickstart.mts
   ```

   **Python** (`quickstart.py`)

   ```python
   import os
   from scalekit import ScalekitClient
   from dotenv import load_dotenv
   load_dotenv()

   scalekit_client = ScalekitClient(
       env_url=os.getenv("SCALEKIT_ENVIRONMENT_URL"),
       client_id=os.getenv("SCALEKIT_CLIENT_ID"),
       client_secret=os.getenv("SCALEKIT_CLIENT_SECRET"),
   )
   actions = scalekit_client.actions

   connection_name = "llmwhisperer"
   identifier = "user_123"

   # Generate an authorization link for the user
   link_response = actions.get_authorization_link(
       connection_name=connection_name,
       identifier=identifier,
   )
   print("Authorize LLMWhisperer:", link_response.link)
   input("Press Enter after authorizing...")

   # Make your first call
   result = actions.execute_tool(
       tool_input={},
       tool_name="llmwhisperer_usage_info_get",
       connection_name=connection_name,
       identifier=identifier,
   )
   print(result)
   ```

   ```bash
   python quickstart.py
   ```

   Each user opens the link once and enters their LLMWhisperer credentials there. See [Authorize a user](/agentkit/tools/authorize/) for the full flow and statuses.

## Tools

Pass the exact name to `execute_tool`, with an input like each tool's example. To ask for a tool that's missing, use the [request form](https://scalekitsupport.portal.usepylon.com/forms/request-a-connector-tool).

**Node.js**

```typescript
const result = await actions.executeTool({
  toolName: 'llmwhisperer_detail_get',
  toolInput: {
    whisper_hash: '<whisper_hash>',
  },
  connector: 'llmwhisperer',
  identifier: 'user_123',
})
```

**Python**

```python
result = actions.execute_tool(
    tool_name="llmwhisperer_detail_get",
    tool_input={
        "whisper_hash": "<whisper_hash>",
    },
    connection_name="llmwhisperer",
    identifier="user_123",
)
```

`result.data` is the app's response as JSON, and `result.execution_id` (`executionId` in Node.js) is the ID of the call.

### `llmwhisperer_detail_get`

Get Extraction Details · Read-only

Get the details and metadata of an extraction job.
Returns mode, requested, processed and total pages, file size, tag, processing time and timestamps (plus token_usage for document insights jobs).
Use this to audit a job; it does not return the text.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `whisper_hash` (`string`, required): whisper_hash returned by llmwhisperer_extract when the job was accepted. Example: a1b2c3d4|0f9e8d7c6b5a49382716a5b4c3d2e1f0.

Example input: `{"whisper_hash":"<whisper_hash>"}`

### `llmwhisperer_document_insights_retrieve`

Retrieve Document Insights · Read-only

Retrieve the handwritten signatures and table of contents found by a Document Insights job. The result can be retrieved only once.
Returns signature_metadata keyed by page, a nested table_of_contents and whisper_metadata with token usage.
Use this for jobs submitted to LLMWhisperer's Document Insights API; check llmwhisperer_status_get first.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `whisper_hash` (`string`, required): whisper_hash returned when the Document Insights job was submitted. Example: a1b2c3d4|0f9e8d7c6b5a49382716a5b4c3d2e1f0.

Example input: `{"whisper_hash":"<whisper_hash>"}`

### `llmwhisperer_highlights_get`

Get Line Highlights · Read-only

Get the bounding-box metadata of extracted lines, used to highlight text in the original document.
Returns, per line number, the page, base_y, height and page_height (absolute and percent).
Works only for jobs started with add_line_nos true.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `lines` (`string`, required): Lines to return, for example 1-5,7,21- for lines 1 to 5, 7 and 21 to the end. Example: 1-20.
- `whisper_hash` (`string`, required): whisper_hash returned by llmwhisperer_extract when the job was accepted. Example: a1b2c3d4|0f9e8d7c6b5a49382716a5b4c3d2e1f0.

Example input: `{"lines":"<lines>","whisper_hash":"<whisper_hash>"}`

### `llmwhisperer_pdf_to_images_status_get`

Get PDF To Images Status · Read-only

Get the status of a PDF-to-images conversion job.
Returns status (accepted, processing, processed, error or retrieved), per-page progress, and page counts once processed.
Use this for jobs submitted to LLMWhisperer's PDF to Images API. The ZIP download itself is not available through this connector.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `whisper_hash` (`string`, required): whisper_hash returned when the PDF to Images job was submitted. Example: a1b2c3d4|0f9e8d7c6b5a49382716a5b4c3d2e1f0.

Example input: `{"whisper_hash":"<whisper_hash>"}`

### `llmwhisperer_status_get`

Get Extraction Status · Read-only

Get the status of a text extraction job.
Returns status (accepted, processing, processed, error or retrieved), a message and per-page detail.
Poll this after llmwhisperer_extract. When status is processed, call llmwhisperer_text_retrieve.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `whisper_hash` (`string`, required): whisper_hash returned by llmwhisperer_extract when the job was accepted. Example: a1b2c3d4|0f9e8d7c6b5a49382716a5b4c3d2e1f0.

Example input: `{"whisper_hash":"<whisper_hash>"}`

### `llmwhisperer_text_retrieve`

Retrieve Extracted Text · Read-only

Retrieve the extracted text of a finished job. The text can be retrieved only once, so store it.
Returns result_text plus confidence_metadata and metadata, or only the text when text_only is true.
Call this once llmwhisperer_status_get reports processed. A second call returns an error.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `whisper_hash` (`string`, required): whisper_hash returned by llmwhisperer_extract when the job was accepted. Example: a1b2c3d4|0f9e8d7c6b5a49382716a5b4c3d2e1f0.
- `text_only` (`boolean`, optional): true to return only the text, false to include the metadata. Defaults to false. Example: false.

Example input: `{"whisper_hash":"<whisper_hash>"}`

### `llmwhisperer_usage_info_get`

Get Usage Info · Read-only

Get the page usage and quotas of your LLMWhisperer account.
Returns subscription_plan, monthly and daily quotas, page counts per mode, overage and today's page count (-1 means not applicable).
Use this to check remaining quota. Use llmwhisperer_usage_stats_get for usage by tag.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs: none.

Example input: `{}`

### `llmwhisperer_usage_stats_get`

Get Usage By Tag · Read-only

Get pages processed per mode for one audit tag over a date range (last 30 days by default).
Returns start_date, end_date, tag and usage (pages_processed per service_type).
Use the tag passed to llmwhisperer_extract. Use llmwhisperer_usage_info_get for account totals.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `tag` (`string`, required): Audit tag to report on. Example: invoices.
- `from_date` (`string`, optional): Start date, YYYY-MM-DD. Example: 2026-10-01.
- `to_date` (`string`, optional): End date, YYYY-MM-DD. Example: 2026-10-31.

Example input: `{"tag":"<tag>"}`

### `llmwhisperer_webhook_get`

Get Webhook · Read-only

Get a registered webhook by name.
Returns the webhook's url, auth_token and webhook_name.
Use this to check a webhook before passing its name as use_webhook.

Read only: yes. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `webhook_name` (`string`, required): Name the webhook was registered under. Example: extraction-done.

Example input: `{"webhook_name":"<webhook_name>"}`

### `llmwhisperer_extract`

Extract Text From Document URL · Write

Start converting a document (PDF, scanned image, Office file or spreadsheet) at a public URL into LLM-ready text, asynchronously.
Returns a whisper_hash and status processing; the text is not in this response.
Poll llmwhisperer_status_get until status is processed, then call llmwhisperer_text_retrieve once. Pass use_webhook to be called back instead of polling.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `url` (`string`, required): Public URL of the document to process. It must be downloadable without authentication. Example: https://example.com/files/invoice.pdf.
- `add_line_nos` (`boolean`, optional): Add line numbers to the text and save line positions for llmwhisperer_highlights_get. Defaults to false. Example: true.
- `allow_rotated_text` (`boolean`, optional): Keep rotated text such as watermarks. form, high_quality and table modes only. Defaults to true. Example: true.
- `file_name` (`string`, optional): Audit file name stored with the job in usage reports. Example: invoice-2026-001.pdf.
- `gaussian_blur_radius` (`integer`, optional): Gaussian blur radius for noise removal, low_cost mode only. Defaults to 0. Example: 0.
- `horizontal_stretch_factor` (`number`, optional): Horizontal stretch to apply, for example 1.1 for 10 percent, when columns merge. Defaults to 1.0. Example: 1.1.
- `include_cell_borders` (`boolean`, optional): For Excel files: draw the workbook's cell borders as box-drawing characters. Defaults to true. Example: false.
- `include_line_confidence` (`boolean`, optional): Add a confidence value per line in the line metadata. Needs add_line_nos true and layout_preserving. Defaults to false. Example: false.
- `line_splitter_strategy` (`string`, optional): Advanced line-splitting strategy. Defaults to left-priority. Example: left-priority.
- `line_splitter_tolerance` (`number`, optional): Fraction of average character height above or below the baseline at which text moves to the next line. Defaults to 0.4. Example: 0.4.
- `mark_horizontal_lines` (`boolean`, optional): Reproduce horizontal lines. Needs mark_vertical_lines true; not for native_text. Defaults to false. Example: false.
- `mark_vertical_lines` (`boolean`, optional): Reproduce vertical lines from the document. Not for native_text. Defaults to false. Example: false.
- `median_filter_size` (`integer`, optional): Median filter size for noise removal, low_cost mode only. Defaults to 0. Example: 0.
- `mode` (`string`, optional): Processing mode. native_text for digital PDFs and Office files, low_cost for clean scans, high_quality for handwriting or poor scans, form for forms and checkboxes, table for dense tables. LLMWhisperer defaults to form. Example: form. One of: `native_text`, `low_cost`, `high_quality`, `form`, `table`.
- `output_mode` (`string`, optional): layout_preserving keeps the document layout for LLMs; text returns plain text. LLMWhisperer defaults to layout_preserving. Example: layout_preserving. One of: `layout_preserving`, `text`.
- `page_separator` (`string`, optional): String placed between pages in the output. LLMWhisperer defaults to <<<. Example: <<<.
- `pages_to_extract` (`string`, optional): Pages to extract, for example 1-5,7,21- for pages 1 to 5, 7 and 21 to the end. All pages by default. Example: 1-5,7.
- `tag` (`string`, optional): Audit tag stored with the job, used to filter llmwhisperer_usage_stats_get. Defaults to default. Example: invoices.
- `use_webhook` (`string`, optional): Name of a webhook registered with llmwhisperer_webhook_create to call when extraction finishes. Example: extraction-done.
- `webhook_metadata` (`string`, optional): Text passed through verbatim to the webhook. Example: order-4471.
- `word_confidence_threshold` (`number`, optional): Minimum OCR confidence, 0 to 1, for a word to be kept. form, high_quality and table modes only. Defaults to 0.3. Example: 0.3.

Example input: `{"url":"<url>"}`

### `llmwhisperer_webhook_create`

Register Webhook · Write

Register a webhook that LLMWhisperer calls with the extracted text when a job finishes. A test payload is sent first; the URL must answer 200.
Returns a confirmation message.
Then pass webhook_name as use_webhook in llmwhisperer_extract.

Read only: no. Destructive: no. Idempotent (safe to retry): no.

Inputs:

- `auth_token` (`string`, required): Bearer token LLMWhisperer sends to your endpoint. Use an empty string if the endpoint has no authentication. Example: my-webhook-secret.
- `url` (`string`, required): URL LLMWhisperer POSTs the result to. It must return 200 to the test payload sent on registration. Example: https://example.com/hooks/llmwhisperer.
- `webhook_name` (`string`, required): Name of the webhook. Pass it as use_webhook when calling llmwhisperer_extract. Example: extraction-done.

Example input: `{"auth_token":"<auth_token>","url":"<url>","webhook_name":"<webhook_name>"}`

### `llmwhisperer_webhook_update`

Update Webhook · Write

Change the URL or auth token of an existing webhook, identified by name.
Returns a confirmation message.
Use llmwhisperer_webhook_create for a new webhook.

Read only: no. Destructive: no. Idempotent (safe to retry): yes.

Inputs:

- `auth_token` (`string`, required): Bearer token LLMWhisperer sends to your endpoint. Use an empty string if the endpoint has no authentication. Example: my-webhook-secret.
- `url` (`string`, required): URL LLMWhisperer POSTs the result to. It must return 200 to the test payload sent on registration. Example: https://example.com/hooks/llmwhisperer.
- `webhook_name` (`string`, required): Name of the webhook. Pass it as use_webhook when calling llmwhisperer_extract. Example: extraction-done.

Example input: `{"auth_token":"<auth_token>","url":"<url>","webhook_name":"<webhook_name>"}`

### `llmwhisperer_webhook_delete`

Delete Webhook · Destructive

Delete a registered webhook by name.
Returns a confirmation message.
Jobs that name this webhook in use_webhook will no longer call it.

Read only: no. Destructive: yes. Idempotent (safe to retry): yes.

Inputs:

- `webhook_name` (`string`, required): Name the webhook was registered under. Example: extraction-done.

Example input: `{"webhook_name":"<webhook_name>"}`

## Related

Other AI connectors ([all 216](/agentkit/connectors/?category=ai)):

- [Chorus](/agentkit/connectors/chorus/): Scalekit connector, Username and password, 9 tools
- [Gong](/agentkit/connectors/gong/): Scalekit connector, OAuth, 69 tools
- [Jiminny](/agentkit/connectors/jiminny/): Scalekit connector, Access token, 27 tools


---

## More Scalekit documentation

| Resource | What it contains | When to use it |
|----------|-----------------|----------------|
| [/llms.txt](/llms.txt) | Structured index with routing hints per product area | Start here — find which documentation set covers your topic before loading full content |
| [/llms-full.txt](/llms-full.txt) | Complete documentation for all Scalekit products in one file | Use when you need exhaustive context across multiple products or when the topic spans several areas |
| [sitemap-0.xml](https://docs.scalekit.com/sitemap-0.xml) | Full URL list of every documentation page | Use to discover specific page URLs you can fetch for targeted, page-level answers |
