diff --git a/CHANGELOG.md b/CHANGELOG.md
index 0ea2591cac62..9134f8cf65c7 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -33,6 +33,7 @@ Docs: https://docs.openclaw.ai
### Fixes
+- **Web Readability relative links:** seed parsed documents with the request URL so article links resolve correctly while removing the plugin's duplicate lazy-loader facade. (#106860)
- **Browser auto-routing:** fall back to the Gateway host when an implicitly selected browser node reports that its control host is unreachable, while preserving explicit node pins and ambiguous action failures.
- **Discord voice participant context:** maintain the live Gateway voice-state roster and include current channel participants in authorized voice agent turns so agents can answer who is present.
- **OC Path JSONC insertion:** patch object and array insertions through `jsonc-parser` so comments, trailing commas, and CRLF formatting survive. (#106847)
diff --git a/extensions/web-readability/web-content-extractor.test.ts b/extensions/web-readability/web-content-extractor.test.ts
index 9a3a0a41f3c8..2a8e32de55b0 100644
--- a/extensions/web-readability/web-content-extractor.test.ts
+++ b/extensions/web-readability/web-content-extractor.test.ts
@@ -20,6 +20,7 @@ const SAMPLE_HTML = `
Example Article
Main content starts here with enough words to satisfy readability.
Second paragraph for a bit more signal.
+ Continue reading
@@ -59,6 +60,7 @@ describe("web readability extractor", () => {
});
const extracted = requireReadabilityResult(result);
expect(extracted.text).toContain("Main content starts here");
+ expect(extracted.text).toContain("[Continue reading](https://example.com/next)");
expect(extracted.title).toBe("Example Article");
});
diff --git a/extensions/web-readability/web-content-extractor.ts b/extensions/web-readability/web-content-extractor.ts
index 09b257d4c5e6..7126ec7253b1 100644
--- a/extensions/web-readability/web-content-extractor.ts
+++ b/extensions/web-readability/web-content-extractor.ts
@@ -1,4 +1,5 @@
// Web Readability plugin module implements web content extractor behavior.
+import { createLazyRuntimeSurface } from "openclaw/plugin-sdk/lazy-runtime";
import type {
WebContentExtractionRequest,
WebContentExtractionResult,
@@ -30,65 +31,20 @@ const HTML_VOID_TAGS = new Set([
"wbr",
]);
-type ParsedHtml = {
- document: Document;
-};
-
-type ParseHtml = (html: string) => ParsedHtml;
-
-type ReadabilityResult = {
- content?: string;
- textContent?: string | null;
- title?: string | null;
-};
-
-type ReadabilityInstance = {
- parse(): ReadabilityResult | null;
-};
-
-type ReadabilityConstructor = new (
- document: Document,
- options: { charThreshold: number },
-) => ReadabilityInstance;
-
-type ReadabilityModule = {
- Readability: ReadabilityConstructor;
-};
-
-type LinkedomModule = {
- parseHTML: ParseHtml;
-};
-
const READABILITY_MODULE = "@mozilla/readability";
const LINKEDOM_MODULE = "linkedom";
-let readabilityDepsPromise:
- | Promise<{
- Readability: ReadabilityConstructor;
- parseHTML: ParseHtml;
- }>
- | undefined;
-
-async function loadReadabilityDeps(): Promise<{
- Readability: ReadabilityConstructor;
- parseHTML: ParseHtml;
-}> {
- if (!readabilityDepsPromise) {
- readabilityDepsPromise = Promise.all([
- import(READABILITY_MODULE) as Promise,
- import(LINKEDOM_MODULE) as Promise,
- ]).then(([readability, linkedom]) => ({
- Readability: readability.Readability,
- parseHTML: linkedom.parseHTML,
- }));
- }
- try {
- return await readabilityDepsPromise;
- } catch (error) {
- readabilityDepsPromise = undefined;
- throw error;
- }
-}
+const loadReadabilityDeps = createLazyRuntimeSurface(
+ () =>
+ Promise.all([
+ import(READABILITY_MODULE) as Promise,
+ import(LINKEDOM_MODULE) as Promise,
+ ]),
+ ([readability, linkedom]) => ({
+ Readability: readability.Readability,
+ parseHTML: linkedom.parseHTML,
+ }),
+);
function exceedsEstimatedHtmlNestingDepth(html: string, maxDepth: number): boolean {
let depth = 0;
@@ -175,12 +131,7 @@ async function extractWithReadability(
}
try {
const { Readability, parseHTML } = await loadReadabilityDeps();
- const { document } = parseHTML(cleanHtml);
- try {
- (document as { baseURI?: string }).baseURI = request.url;
- } catch {
- // Best-effort base URI for relative links.
- }
+ const { document } = parseHTML(cleanHtml, { location: { href: request.url } });
const reader = new Readability(document, { charThreshold: 0 });
const parsed = reader.parse();
if (!parsed?.content) {