As measured by Artificial Analysis, time to final transcription, for example, improves by 70%. On the FLEURS benchmark across a set of top languages and locales, the model delivers precise multilingual performance, improving over Chirp 3, and achieving a 5.50% WER in streaming mode and 5.04% WER in non-streaming use-cases.

Complete reference for the tools Claude Code can use, including permission requirements and per-tool behavior.

We did not enable internet access or inter-agent communication for many of the environments in these training experiments. Despite these restrictions, the agents discovered ways to exploit our research infrastructure to communicate with one another and access the internet.

The models first found ways to communicate by writing files into the Artifactory package manager. This effectively turned Artifactory into an unintended message board, where agents could exchange information with one another. Read more: How agents used the message board 

The agents also found ways to get access to the internet. Using an exploit⁠(opens in a new window), the agents made Artifactory, which itself had access to the internet for the limited purpose of downloading and installing packages, send any internet requests on their behalf. After the agents discovered these methods for obtaining internet access, they also shared them with other agents through the message board, which provided information to enable additional agents to exploit our infrastructure.

Claude Code also keeps a fifth file, ~/.claude.json, that it writes for itself; you don’t need to edit it. It holds your sign-in session, MCP server configurations, per-project state such as trust decisions, and the global config keys that /config writes for you.

Claude Code stores its configuration and auth credentials in ~/.claude by default. The CLAUDE_CONFIG_DIR environment variable overrides this location. By creating a separate config directory per account - each with its own OAuth credentials - and symlinking shared resources back to a canonical source, you get:

• Separate billing/authentication per account
• Shared sessions - resume any session from either account
• Shared settings, plugins, commands, plans, and skills - configure once, use everywhere

​ One-Time Setup This setup is done once on your machine, outside of Maestro. ​

1. Authenticate Each Account Start Claude Code normally and complete OAuth for your first account:

claude

Complete OAuth for account A (e.g., personal)

Copy the authenticated config to a named directory:

cp -a ~/.claude ~/.claude-personal

Then authenticate your second account:

mv ~/.claude/.claude.json ~/.claude/.claude.json.bak claude

Complete OAuth for account B (e.g., work)

cp -a ~/.claude ~/.claude-work rm ~/.claude/.claude.json.bak

The main ~/.claude/ directory doesn’t need its own .claude.json. It serves as the canonical source for shared resources.

​ 2. Symlink Shared Resources For each account directory, replace local copies with symlinks back to ~/.claude so settings, plugins, and sessions stay in sync:

Repeat for each account directory (e.g., ~/.claude-personal, ~/.claude-work)

CONFIG_DIR=~/.claude-personal

Back up directories that will be symlinked

mv $CONFIG_DIR/projects $CONFIG_DIR/projects-pre mv $CONFIG_DIR/todos $CONFIG_DIR/todos-pre mv $CONFIG_DIR/session-env $CONFIG_DIR/session-env-pre

Remove files/dirs that will become symlinks

rm -rf $CONFIG_DIR/commands $CONFIG_DIR/ide $CONFIG_DIR/plans $CONFIG_DIR/plugins $CONFIG_DIR/skills rm -f $CONFIG_DIR/settings.json $CONFIG_DIR/CLAUDE.md

Create symlinks

ln -s ~/.claude/commands $CONFIG_DIR/commands ln -s ~/.claude/ide $CONFIG_DIR/ide ln -s ~/.claude/plans $CONFIG_DIR/plans ln -s ~/.claude/plugins $CONFIG_DIR/plugins ln -s ~/.claude/skills $CONFIG_DIR/skills ln -s ~/.claude/settings.json $CONFIG_DIR/settings.json ln -s ~/.claude/CLAUDE.md $CONFIG_DIR/CLAUDE.md ln -s ~/.claude/todos $CONFIG_DIR/todos ln -s ~/.claude/session-env $CONFIG_DIR/session-env ln -s ../.claude/projects $CONFIG_DIR/projects

Set CLAUDE_CONFIG_DIR to your account’s config path (e.g., /Users/you/.claude-personal)

How is Social Security funded?

The Old-Age and Survivors Insurance (OASI) and Disability Insurance (DI) trust funds, which were worth a combined $2.56 trillion at the end of 2025, fund Social Security.

US workers pay into the trust funds through payroll taxes and employers match these contributions. Self-employed people are taxed at twice the rate — essentially paying as the employer and as the employee. The SSA estimates that workers contributed $1.32 trillion to the trust funds in 2025. The funds also generated $58 billion from income taxes on benefits and $69 billion on interest.

In June 2026, the SSA projected that the trust funds combined reserves—OASI and DI—could be depleted by 2034, barring any intervention from Congress, at which point only about 83% of scheduled benefits could be covered by incoming revenue. They further projected that OASI could face depletion in the fourth quarter of 2032.

The nation’s population is aging, posing a challenge for the Social Security system, which is designed for — among other purposes — its workers to support its retirees. The model relies on some level of balance between workers paying into the trust funds and beneficiaries receiving those funds.

In recent years, though, that balance has been impacted by three factors in particular:

• More people reaching retirement age
• Relatively fewer entering the workforce
• People living longer into retirement

We charge a simple, straightforward fee depending on the type of exchange you are doing. We are so confident in our specialized approach that our Standard Exchange services are offered with a "pay us when you close" guarantee – if you decide not to go through with your exchange for any reason, you’ll get your money back.

$100 Million Fidelity Bond

Protects the taxpayer in the unlikely event of a company failure; the bond is there to satisfy the financial obligations of the company.

$50 Million Written Performance Guaranty

Ensures the taxpayer exchange funds are returned in the amount of deposit.

$30 Million Errors & Omissions Insurance

Safeguards against employee negligence resulting in a loss to the taxpayer

IPX1031’s parent company, Fidelity National Financial (FNF) is the world’s largest provider of title insurance and real estate related services.

Inboxes, calendars, newsletters, bookings, an AI assistant, and an API for your agents. Everything on your own domain, in one subscription. From $4/mo.

hello@, orders@, support@, across every domain you own. Read and reply from one place, and every reply leaves from the right address.

Squire is an HTML5 rich text editor, which provides powerful cross-browser normalisation in a flexible lightweight package (only 16KB of JS after minification and gzip, with no dependencies!).

It was designed to handle email composition for the Fastmail web app. The most important consequence of this (and where Squire differs from most other modern rich text editors) is that it must handle arbitrary HTML, because it may be used to forward or quote emails from third-parties and must be able to preserve their HTML without breaking the formatting. This means that it can't use a more structured (but limited) internal data model (as most other modern HTML editors do) and the HTML remains the source-of-truth. The other consequence is excellent handling of multiple levels of blockquotes.

Fast, private email hosting for you or your business.

Amazing webmail+calendar+contact+files client.

It is possible to use an NFS file system for storing the PostgreSQL data directory. PostgreSQL does nothing special for NFS file systems, meaning it assumes NFS behaves exactly like locally-connected drives. PostgreSQL does not use any functionality that is known to have nonstandard behavior on NFS, such as file locking.

The only firm requirement for using NFS with PostgreSQL is that the file system is mounted using the hard option. With the hard option, processes can “hang” indefinitely if there are network problems, so this configuration will require a careful monitoring setup.

It is not necessary to use the sync mount option. The behavior of the async option is sufficient, since PostgreSQL issues fsync calls at appropriate times to flush the write caches. (This is analogous to how it works on a local file system.) However, it is strongly recommended to use the sync export option on the NFS server on systems where it exists (mainly Linux). Otherwise, an fsync or equivalent on the NFS client is not actually guaranteed to reach permanent storage on the server, which could cause corruption similar to running with the parameter fsync off.

After watching this amazing talk, I realized that as long as we can guarantee that WAL (Write-Ahead Logs) buffers are written to the network storage, the database should be able to recover from the errors in case of a failure. Armed with this knowledge, I made an attempt on storing Postgres data on an NFS share.

A library that provides an embeddable, persistent key-value store for fast storage.

This code is a library that forms the core building block for a fast key-value server, especially suited for storing data on flash drives. It has a Log-Structured-Merge-Database (LSM) design with flexible tradeoffs between Write-Amplification-Factor (WAF), Read-Amplification-Factor (RAF) and Space-Amplification-Factor (SAF). It has multi-threaded compactions, making it especially suitable for storing multiple terabytes of data in a single database.

Existing Gmail and Outlook MCP servers route queries through Google/Microsoft APIs. With stalwart-mcp, AI queries hit your Stalwart instance directly. Mail data never leaves your infrastructure; only the AI questions and answers go to the LLM provider.

JMAP is the open protocol that replaces IMAP (and CardDAV and CalDAV). It’s faster, simpler, more powerful, and built for how email works today.