@crawlora-org/sdk 1.2.0-sdk.15 → 1.2.0-sdk.16

package/CHANGELOG.md

@@ -1,5 +1,12 @@
 # Changelog
 
+## v1.2.0-sdk.16
+
+- Documented response headers on `CrawloraError`, case-insensitive header
+  overrides, strict response modes, `Retry-After` retries, and abort semantics.
+- Added docs coverage checks for the release-polish behavior.
+- Kept the generated operation contract unchanged.
+
 ## v1.2.0-sdk.15
 
 - Added case-insensitive request header overrides across auth, user-agent, and

package/README.md

@@ -87,7 +87,9 @@ const crawlora = new CrawloraClient({
 ```
 
 Per-request options can override headers, timeout, abort signal, and response
-mode:
+mode. Header names are matched case-insensitively, so request headers can
+override default auth, user-agent, and content headers without duplicating
+variants such as `x-api-key` and `X-API-KEY`:
 
 ```js
 const controller = new AbortController();
@@ -104,8 +106,9 @@ const response = await crawlora.bing.search(
 
 ## Text Responses
 
-Most endpoints return JSON. Endpoints that support alternate text output, such
-as YouTube transcripts, can opt into text mode:
+Most endpoints return JSON. `responseType` must be `auto`, `json`, or `text`.
+Endpoints that support alternate text output, such as YouTube transcripts, can
+opt into text mode:
 
 ```js
 const transcript = await crawlora.youtube.transcript(
@@ -133,8 +136,11 @@ try {
 }
 ```
 
-The error includes `status`, optional API `code`, parsed `body`, the original
-`response` when available, and a parser or transport `cause` when relevant.
+The error includes `status`, optional API `code`, parsed `body`, response
+`headers`, the original `response` when available, and a parser or transport
+`cause` when relevant. Retryable responses honor positive `Retry-After`
+headers, capped at 30 seconds. Externally aborted requests fail with
+`Crawlora request aborted` and are not retried.
 
 ## Examples
 

package/docs/recipes.md

@@ -47,8 +47,14 @@ const result = await crawlora.bing.search(
 );
 ```
 
+Request headers override default auth, user-agent, and content headers
+case-insensitively. Retryable API responses honor positive `Retry-After`
+headers, capped at 30 seconds. Externally aborted requests are not retried.
+
 ## Text Responses
 
+`responseType` must be `auto`, `json`, or `text`.
+
 ```js
 const text = await crawlora.youtube.transcript(
   { id: "VIDEO_ID", format: "text" },
@@ -63,7 +69,7 @@ try {
   await crawlora.bing.search({ q: "coffee shops" });
 } catch (error) {
   if (error instanceof CrawloraError) {
-    console.log(error.status, error.code, error.body);
+    console.log(error.status, error.code, error.body, error.headers);
   }
 }
 ```

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@crawlora-org/sdk",
-  "version": "1.2.0-sdk.15",
+  "version": "1.2.0-sdk.16",
   "description": "JavaScript and TypeScript SDK for the public Crawlora API.",
   "keywords": [
     "crawlora",

package/src/client.js

@@ -1,7 +1,7 @@
 import { operations, groups } from "./operations.js";
 
 const DEFAULT_BASE_URL = "https://api.crawlora.net/api/v1";
-export const VERSION = "1.2.0-sdk.15";
+export const VERSION = "1.2.0-sdk.16";
 const DEFAULT_USER_AGENT = `crawlora-js-sdk/${VERSION}`;
 
 export class CrawloraError extends Error {