Skip to content
DailyPrompt

your daily dose of claude

01 product

Claude Gains Connectors for Everyday Life: Spotify, Instacart, Booking.com and More

Anthropic has launched a new wave of personal app connectors that let Claude work directly with services like Spotify, Instacart, Tripadvisor, Booking.com, AllTrails, Audible, Uber, TurboTax, and more. Users can now chain real-world tasks across services in a single conversation — plan a trip, book a table, order groceries, and queue a playlist without switching apps. Connectors are available on Claude.ai, the desktop app, and mobile (iOS/Android), with connected-app data not used to train models.

Claude Blog →
product

Built-in Memory for Claude Managed Agents Arrives in Public Beta

Anthropic has shipped memory support for Claude Managed Agents, allowing agents to learn and improve across sessions rather than starting fresh each time. Memories are stored as exportable files, letting developers control exactly what agents retain and share, with full API management and audit logs. Early adopters including Rakuten and Wisedocs reported significant gains in accuracy, cost, and latency after enabling the feature.

Claude Blog
C
Claude
@claudeai

Claude can now connect to more of the apps you use outside of work, including @Tripadvisor, @bookingcom, @resy, @Instacart, @Spotify, @audible_com, @AllTrails, @thumbtack, intuit @turbotax, and more.

11h View on X →
announcement

$ cat story_03.md

Anthropic Admits Three Bugs Quietly Degraded Claude for Weeks, Resets All Usage Limits

After weeks of user complaints, Anthropic published a detailed postmortem revealing three separate changes between March and April caused Claude to feel slower, more forgetful, and less capable. The culprits were a reasoning-effort downgrade, a caching bug that erased thinking history every turn, and an overly strict verbosity prompt — all now fixed in v2.1.116 or earlier. Anthropic has reset usage limits for all subscribers and pledged stricter internal testing to prevent repeat incidents.

$ open anthropic-engineering →
announcement No. 04

Anthropic and NEC Team Up to Build Japan's Largest AI Engineering Workforce

NEC Corporation has become Anthropic's first Japan-based global partner, deploying Claude across 30,000 employees worldwide and building what it calls Japan's largest AI-native engineering team through a new internal Centre of Excellence. The partnership will produce industry-specific AI products for finance, manufacturing, and local government, with Claude Opus 4.7 integrated into NEC's BluStellar platform and cybersecurity operations. Joint development is already underway as part of NEC's 'Client Zero' programme.

Source: Anthropic →
C
Claude
@claudeai

Memory on Claude Managed Agents is now in public beta. Your agents can now learn from every session, using an intelligence-optimized memory layer that balances performance with flexibility.

8h View on X →
05
community

Security Researcher Flags Claude Desktop for Silently Installing Browser Bridges on macOS

A security researcher discovered that Claude Desktop automatically installs a Native Messaging manifest into multiple Chromium browser profiles on macOS — including for browsers not yet installed — pre-authorising three extension IDs without any user prompt. Anthropic acknowledged the issue, clarified that no data is sent to its servers without explicit user action, and said patch version 1.2.4 is on the way to ensure complete removal on uninstall. The episode has reignited debates about silent permission changes in desktop AI apps.

Malwarebytes →
C
ClaudeDevs
@ClaudeDevs

Over the past month, some of you reported Claude Code's quality had slipped. We investigated, and published a post-mortem on the three issues we found. All are fixed in v2.1.116+ and we've reset usage limits for all subscribers.

12h View on X →