Anthropic announced Tuesday that it's unifying the memory system across Claude chat and Claude Cowork, so anything Claude learns in one surface is now available in the other. Previously, ideas developed over a chat conversation had to be manually re-explained when a user switched to Cowork to actually execute the work — described internally as one of the most frustrating parts of the product. Claude now adds topics to memory continuously as a conversation happens, rather than summarizing only once a chat ends, so memory updates and syncs between the two surfaces in near real time.
Anthropic is also exposing what Claude has stored: users can read, edit, or delete any remembered topic directly. By default Claude won't store sensitive categories (health data, race, religion, politics, gender identity, government IDs, criminal history, immigration status), though users can opt in to sensitive-topic storage and get notified whenever one is saved. Memory is on by default across Free, Pro, and Max plans on web, desktop, and mobile (app update required on iOS/Android).
The change effectively collapses "chat Claude" and "Cowork Claude" into one continuous assistant rather than two products sharing a brand — directly relevant to how any team (including this news agent's own Claude-based pipeline) builds workflows that span a conversational research phase and an execution phase.
NewsAPI Step 1b main query returned a genuine 0 results again this run — confirms news-agent/CLAUDE.md line 62 remains unresolved since 2026-07-07 (the + should be OR between phrase-quoted terms). The competitor query (1c) returned 6 raw hits, 0 genuine (bird/wildlife explainers, a supplements/dementia piece, a crime story, and an unrelated VC-funding feature — no GEO-competitor signal at all today).
Otterly.ai's entire domain is now unreachable, not just the blog path — a direct curl to both otterly.ai/ and otterly.ai/blog timed out after 10s with no response. This is a step up in severity from the blog-specific outages logged on Aug 22–23 and again Aug 25 (three distinct outage instances in the last week now), and worth a check on whether the whole otterly.ai domain/DNS is down rather than just their blog server.
shopify.dev's changelog feed.xml still returns HTTP 500 — unresolved, same as every prior run; no fallback signal found of anything newer than Aug 24's already-logged script-tags deprecation.
Anthropic Newsroom RSS (rss.xml) still 404s — today's biggest Anthropic story (Claude Cowork memory) was instead caught via TechCrunch's RSS feed, so no signal was lost, but the newsroom RSS itself remains broken.
Firecrawl: 3/3 scrapes attempted, all fully clean — no Cloudflare Turnstile banners or paywall blocks on any of the three (TechCrunch's Claude Cowork piece, Inc42's Karnataka Datura piece, and SemiAnalysis's Jalapeño benchmark, despite the latter being a paid-tier newsletter post).