dannykeane

Official TypeScript SDK and CLI for the dannykeane.com public API: Danny Keane's work history, GitHub contribution calendar, live visitor stats, and the Habbo room NPC chat.

Zero dependencies, native fetch. ESM with bundled types, Node 18.3+.

npm i dannykeane

Usage

import { DannyKeane } from "dannykeane";

const dk = new DannyKeane();
for await (const role of dk.iterWork()) console.log(role.company, role.years);

Every endpoint is public, so there is no API key. Pass baseUrl to point at a local or preview deploy.

| Method | API operation | Endpoint | | --- | --- | --- | | listWork({ limit?, cursor? }) | listWork | GET /api/v1/work | | iterWork({ limit? }) | listWork, every page | GET /api/v1/work | | getStats() | getStats | GET /api/v1/stats | | getGithubContributions() | getGithubContributions | GET /api/v1/github-contributions | | npcChat({ message, npcId, context? }) | npcChat | POST /api/v1/npc-chat | | npcChatHealth() | npcChatHealth | GET /api/v1/npc-chat | | runBatch(["getStats", ...]) | runBatch | POST /api/v1/batch |

Every method also takes a last { signal } argument for cancellation. The full contract is the OpenAPI document at https://dannykeane.com/openapi.json.

Pagination

listWork returns one page: { data, next_cursor }, with next_cursor null on the last page. iterWork is an async iterator that follows the cursors for you.

Errors

Every non-2xx response rejects with ApiError, parsed from the API's ErrorResponse body:

import { ApiError, DannyKeane } from "dannykeane";

try {
  await new DannyKeane().listWork({ limit: 50 });
} catch (error) {
  if (error instanceof ApiError) console.log(error.status, error.code, error.message, error.hint);
  // 400 invalid_parameter limit must be an integer from 1 to 20 ...
}

error.body holds the full parsed body. Network failures reject with fetch's own error.

Rate limits

On a 429 the client waits out Retry-After and retries, up to maxRetries (default 2). A Retry-After longer than maxRetryDelay (default 60 seconds) rejects immediately instead, with the wait on error.retryAfter. Set maxRetries: 0 to handle every 429 yourself.

Options

new DannyKeane({
  baseUrl: "http://localhost:3000", // default https://dannykeane.com
  maxRetries: 2,
  maxRetryDelay: 60, // seconds
  timeout: 30_000, // ms, per attempt
  userAgent: "my-agent/1.0", // appended to dannykeane-typescript/<version>
  fetch: customFetch,
});

CLI

The package ships a dannykeane command. Run it without installing:

npx dannykeane work --limit 5

| Command | Operation | | --- | --- | | dannykeane work [--limit N] [--cursor TOKEN] [--all] | listWork; --all walks every page and prints one array | | dannykeane stats | getStats | | dannykeane github-contributions | getGithubContributions | | dannykeane npc-chat [--npc-id ID] <message> | npcChat | | dannykeane npc-chat-health | npcChatHealth | | dannykeane batch <operationId>... | runBatch |

The API's JSON goes to stdout, indented when stdout is a terminal (--pretty forces it). Failures go to stderr as one line of ErrorResponse JSON, with stdout empty. --base-url or DANNYKEANE_BASE_URL points it at another deploy.

Exit codes: 0 success, 1 the API returned an error, 2 usage error, 3 the API could not be reached.

The PyPI package dannykeane ships the same client and CLI for Python.

Links

• Developer docs: https://dannykeane.com/developers
• OpenAPI spec: https://dannykeane.com/openapi.json
• Contact: [email protected]

License

MIT