Content Quality: Compact News item of 758 words (range 400-1200) built around a comparison of InfoQ's November 14 report with Node's release posts, documentation and pull requests. Facts, comparisons and InfoQ's characterisations are kept apart; the Herald's own comparisons are labelled as such and every disagreement is reported with attribution rather than reconciled. Contains no 27.0.0-alpha statement (the one covered by the 2026-11/01 Lithium corrections record) and links the earlier Lithium article with a valid internal /article/ path.
Source Verification: Read the stored snapshots (decompressed) and cross-checked with the GitHub API and the PR/commit text. All 16 snapshots status 200, no archive fallback, suspicious_patterns null for every source (no injection scanner flag); snapshot sha256 of source-15 (InfoQ) and source-11 recomputed and equal to the manifest. source-15 (InfoQ, Daniel Curtis, Nov 14 2026, datePublished 2026-11-14T08:03:00+0000): every InfoQ quote in the article is verbatim ('The additions shipped as part of Node.js 26.10.0', 'essentially every app.', 'one of the most commonly vendored dependencies in the Node.js ecosystem', 'a visible trend in Node.js of reducing reliance on userland utilities', 'further Buffer and http helpers', 'silently broke addon loading from the real filesystem', 'carried'). Grep of the InfoQ body text: no occurrence of '24', 'Krypton', '24.22', 'v24', 'Claude', 'Anthropic' or 'Assisted'; it names 26.10.0 and 26.11.0 (the latter for later Buffer/http helpers) and mentions Bun and Deno once in a comparison, so 'names only 26.10.0' is true for the three features and the article correctly says it does not mention Node.js 24. source-12 (26.10.0 post, 2026-09-22) and source-11 (24.22.0 'Krypton' LTS post, 2026-10-14): Notable Changes in both list crypto.parsePKCS12() #65627, util.markPromiseAsHandled #65805, util.throttle and util.debounce #65899; only 26.10.0 lists 'ffi: load libraries from a mounted VFS (Matteo Collina) #65909'; 24.22.0 has no ffi/vfs entry. source-9, source-6 (v26.x and v24.x util pages): 'Added in: v26.10.0' / 'Added in: v24.22.0' for util.debounce, util.throttle, util.markPromiseAsHandled; source-7, source-5 (crypto pages): parsePKCS12 'Added in: v26.10.0' / 'v24.22.0'; option lists (leading, rejectOnCancel, signal; concurrency, maxPending, overflow, signal, strict; ERR_THROTTLED; privateKey/certificate/additionalCertificates) match the util/crypto text. source-4 (v22.23.5 util page): zero 'debounce' matches, article scopes the claim to that page. source-8, source-10: 'Stability: 1 - Experimental'. source-13 (26.11.0, 2026-10-07): isLatin1, Buffer.stringLength(), isValidHeaderName()/isValidHeaderValue() in Notable Changes. source-14 (26.13.0 'Lithium' LTS, 2026-10-28): LTS transition confirmed. source-2 / GitHub API #65899 ('util: implement debounce', jasnell, landed 2026-09-15): mcollina review 'lgtm I use these in essentially every app.'; ljharb asked 'what's the userland precedent for an AbortSignal-aware debounce implementation?'; the objection that AbortSignals can only transition once is bakkot's comment ('The use of AbortSignal here is a bit weird, and in my opinion wrong'); jasnell 'Updated to make it so that the signal causes all future calls to fail and added the leading option'; bricss asked for throttle. Files changed in #65899 contain no markPromiseAsHandled code; source-1 / API #65805 'src,lib: add util.markPromiseAsHandled' is a separate PR (opened 2026-09-05, merged 2026-09-07, before #65899 was opened on 2026-09-08), so InfoQ's 'same change' differs from Node's own listing; the article reports both with attribution and does not reconcile them (policy rule 6). source-3 / API #65909 (ffi VFS, Collina, co-authored Philipp Dunkel): PR text and commit message say 'loading any addon from the real file system failed with EINVAL while a VFS was mounted', i.e. the bug occurred only while a VFS was mounted, which InfoQ's wording omits; the commit message is a third place that gives the same qualifier. source-0 / API #65627 (crypto.parsePKCS12, bmuenzenmeyer, merged 2026-09-12): body reads 'I (with Claude) noticed Node.js already parses this internally', 'This is my first significant contribution to core that touches the internals' and 'Assisted-by: Claude Opus' - the article's quotes ('with Claude', 'first significant contribution to core that touches the internals') are verbatim and its Claude statements are strictly attributive to the PR page. Allowlist: github.com, nodejs.org, infoq.com all present (chief:review allowlist checks passed); no additions. Independent outlets: 1 (InfoQ); the rest are Node.js first-party pages and GitHub PRs; a search for a second outlet (web search budget exhausted; DuckDuckGo, Bing and Hacker News queries returned nothing) found none, and the article's absence statement is scoped to 'the Herald found'. Anthropic disclosure: required - the article reports, in a passage of its own, a PR that credits 'with Claude' and carries an 'Assisted-by: Claude Opus' line; the one-sentence disclosure is present at the end ('The Herald's contributor model is built by Anthropic; the statements about Claude above are those of the pull request page.') and is neutral. Bun appears only in InfoQ's text, not in the article, so it does not trigger the rule by itself.
Factual Accuracy: Headline verified: all three items are listed in the Notable Changes of both 26.10.0 and 24.22.0 and carry 'Added in' entries for both versions in the docs. Claim-accuracy tests run (quantifier grep over title, summary and body): 'both'/'only'/'each' sentences checked item by item against each source; one imprecision found: the summary says InfoQ 'names only 26.10.0' without the headline's 'for them' scope, while InfoQ also names 26.11.0 (for different helpers) and the body itself says it names 26.10.0 and 26.11.0 - filed as a clarification. Dates: calendar dates throughout, basis stated ('the ones printed on Node's release posts'), reports in chronological order. Derived claims (InfoQ does not mention Node.js 24; 24.22.0 has no ffi entry; v22 page has no debounce) are labelled the Herald's and recomputed from raw text. Differences between InfoQ and the primary pages (markPromiseAsHandled PR, AbortSignal objection, dlopen qualifier) are each shown with the exact attribution above.
Overall Assessment: Accurate, carefully attributed cross-check of an InfoQ report against Node's release posts, docs and pull requests; headline, lead and every subordinate claim verified from raw text, with one scope clarification filed for the summary. Publishable with a corrections record. Featured candidacy: 1 - accurate release coverage for a defined Node.js developer audience, but the features are seven weeks old and the story rests on one outlet