Anthropic
26 articles
-
Claude CodeClaude Code's Sandbox Ran Your Shell Config Outside It
On Wednesday 22 October Claude Code 2.1.309 fixed sandboxed Bash commands running the shell's startup files outside the sandbox before each command. On 2.1.308 every sandboxed command started as an unsandboxed /bin/zsh -c that sourced ~/.zshenv with full access in the project directory, so a .env a sandboxed command wrote was executed with full access on the next Bash call; verified on macOS with an instrumented ZDOTDIR, canary file in the home directory on 2.1.308, denied on 2.1.309 and 2.1.310. The same release stopped sandboxed commands editing the file passed with --mcp-config (appendable on 2.1.308, operation not permitted on 2.1.310). Monday's 2.1.305 added sandbox.allowNestedUserNamespaces, which the docs say turns off a security boundary, and Thursday's 2.1.310 fixed Windows PreToolUse and PermissionRequest hooks letting calls through when Git Bash is missing. The per-session shell snapshot still runs your login shell outside the sandbox at a session's first Bash command, with the project as its working directory, so a leftover .env still runs there on 2.1.310.
-
AnthropicAnthropic's New SDKs Can Drop Your API Key Behind a Gateway
On 21 October Anthropic released API version 2026-10-21, the first since 2023-06-01, with new SDK majors (Python 2.0.0, TypeScript 1.0.0, Go 2.0.0, Java 3.0.0, C# 13.0.0, PHP 1.0.0, Ruby 2.0.0, ant CLI 2.0.0) that send it by default and on which temperature, top_p and top_k return a 400 even at default values. The SDKs also move the API key from X-Api-Key into Authorization: Bearer. On a local capture server, a gateway token set in that header through default_headers replaced the key outright with no error; the guide's fix (auth_token plus X-Api-Key in default_headers, with the gateway stripping its own Authorization) and the undocumented version pin (default_headers anthropic-version 2023-06-01 plus extra_body temperature) both produced the expected request. LiteLLM reads Authorization before x-api-key, so its virtual keys still arrive.
-
Claude CodeClaude Code Compacts Idle Chats, and Has Since September
On 14 October the Claude Code 2.1.299 changelog documented idle compaction: a 200K+ token terminal conversation left idle for about 54 minutes is summarized before its one-hour prompt cache expires. The behavior had been live since 2.1.286 on 30 September through a server-side flag, the idleCompaction opt-out shipped in 2.1.290 on 5 October without a changelog line and was documented by a user reading the binary, and the /config row came in 2.1.302 on 16 October. At Opus 5.5 list prices it trades a $1.60 cold rebuild of a 200,000-token prefix for a $0.04 cache read and a roughly $0.40 summary, runs only on the one-hour TTL that subscribers get within plan usage, summarizes away path-scoped rules, nested CLAUDE.md files and hook-injected context, and records itself as trigger auto, so a PreCompact hook can only tell it from a context-limit compaction by checking the transcript's age.
-
AnthropicHaiku 5.5 Is 10x Cheaper Until Its Tokenizer Catches Up
Claude Haiku 5.5 launched on 7 October at $0.10/$0.50 per million tokens, a tenth of Haiku 4.5, but past 100,000 prompt tokens every rate is five times higher and applies to the whole request, cached tokens included. The newer tokenizer counts about 30% more tokens, so a 77,000-token Haiku 4.5 prompt lands over the line unchanged. It is the only Claude model priced by length; OpenAI's line is at 272K at 2x, Gemini 3.8 Flash has none. Gateways mispriced it, Haiku 4.5's retirement floor ends 15 October, and temperature, prefill and budget_tokens now return 400.
-
AgentsManaged Agents Stopped Fetching Tool Links. Curl Didn't.
Since October 7, web_fetch in Claude Managed Agents only opens URLs that already appeared in the session through a user message or a web search or fetch result, so links found by bash, read or an MCP tool now fail with url_not_in_prior_context. The rule covers one tool, not the sandbox, and the API defaults to unrestricted networking, so bash curl still reaches whatever your environment allows.
-
EngineeringThe Agent That Deleted the Jobs Ran Off Anthropic's Monitor
Anthropic's 17 September oversight numbers: about 30,000 agents at once on its main internal platform, 100% of actions through an online monitor, one decision in 47,000 blocked, and about 50 cases a week reaching a person. That coverage is scoped to one platform, and the August risk report describes unmonitored agents launched with --dangerously-skip-permissions that deleted a large number of jobs. Count coverage by entry point, stamp agent work with agent identity, and track recall, because the filter that makes the review queue manageable cut red-team recall from 94% to 79%.
-
EngineeringClaude's Compaction Summary Comes Back Signed and Sealed
Anthropic's Messages API can now compact a conversation on demand, returning a signed summary block you send in place of the messages it covers. The API rejects an altered block, so anything the summary drops has to be restated after it or rebuilt from the original messages. Failed summaries still return 200, a missing block goes unnoticed, mid-conversation system messages stop applying, and compaction cost only shows up in usage.iterations.
-
AgentsSonnet 5 Costs $2 Today and $3 in September. Model for September.
Claude Sonnet 5 runs agents at near-flagship quality, but the launch price is a promotion that expires August 31 and jumps 50 percent. Model your agent economics on the September number, not the intro rate.
-
AgentsAnthropic Turned Its Coding Agent Into a Scientist
Claude Science is Claude Code pointed at a new toolbox. Same model, same autonomous loop, a reproducibility layer bolted on. The lesson for builders: the harness is the product, and your vertical is next.
-
Agents
Your Agent Logs In With a Password That Never Expires
Most AI agents authenticate with a long-lived static API key in an env var. Anthropic's Workload Identity Federation, GA on June 17, swaps it for short-lived scoped credentials your stack already knows how to issue.
-
AnthropicThe Benchmarks Started Measuring Endurance
On classic SWE-bench the frontier models are bunched within a dozen points. On the long-horizon benchmarks the leader doubles second place. The new tests finally measure what buyers pay for: staying on a messy task for hours.
-
GoogleGoogle Just Bet $40 Billion That Anthropic Wins
Google committed $40 billion to Anthropic, the largest single AI investment ever.
-
AnthropicThe Best AI Model Exists. You Can't Have It.
Anthropic has a model 15 points above Opus on coding benchmarks. You cannot use it. The reason is compute economics.
-
LaborSeven Weeks Later, the Anthropic Labor Data Looks Worse
Seven weeks after Anthropic published their labor data, the picture has worsened. Snap said the quiet part.
-
Anthropic
OpenAI is losing. And it's not close.
1. Anthropic passed OpenAI in revenue. $30 billion annualized run rate, up from $1 billion fourteen months ago. Anthropic wins 70% of enterprise deals in head-to-head competition. That's not a rounding error.
-
Anthropic
Claude Code ultraplan separates thinking from doing
1. Ultraplan moves planning out of the terminal and into a browser. You get inline comments, structured review, and the ability to keep coding while the plan builds itself in the cloud. It sounds minor. It changes how you work.
-
Anthropic
Anthropic's Too-Dangerous Model Escaped Its Sandbox
1. Mythos found zero-days in every major OS and every major browser. Not theoretical weaknesses. Working exploits. Some of these bugs had survived 27 years of human review.
-
Anthropic
What a model tier above Opus actually unlocks
1. The jump from Opus to Capybara isn't incremental. Recursive self-correction changes what you can trust a model to do without babysitting it.
-
Anthropic
14 Rules Claude Follows That Your Team Should Steal
The most useful part of Claude Code's 13,000-token system prompt isn't the identity framing or the tool descriptions. It's a section called "Doing tasks" that contains 14 explicit constraints on how code should be written.
-
Anthropic
28 Prompt Files Inside Claude Code's Agent
Production-grade AI agents don't run on a single system prompt. They run on layered architectures of specialized instructions, each solving a distinct problem, composed at runtime based on context.
-
Anthropic
Claude Code Has a Hidden Pet. It Has Stats.
Somewhere in a TypeScript codebase spanning half a million lines, an Anthropic engineer sat down and drew ASCII art of an axolotl wearing a wizard hat. Then they gave it stats.
-
Anthropic
Things I Learned from the Claude Code Source Code
The Claude Code CLI ships as a compiled binary, but the TypeScript source underneath is remarkably readable once you unpack it. I spent a week going through all 512,000 lines across 1,884 files, looking for the engineering decisions that reveal where
-
AnthropicClaude Code Hit $1 Billion Faster Than Slack or Zoom
Claude Code reached $1 billion in annualized revenue in six months, faster than ChatGPT, Slack, or Zoom. A terminal tool outpaced every enterprise product in history, and the reasons why should worry every SaaS vendor.
-
AnthropicAnthropic Launched an Enterprise Marketplace and Nobody Blinked
Anthropic's new marketplace lets enterprise customers buy third-party Claude apps through existing budget commitments. This is a platform play, not a model update, and it changes the competitive dynamics.
-
Security24,000 Fake Accounts and the New Shape of Industrial Espionage
Three Chinese AI labs created 24,000 fake accounts on Anthropic, generating 16 million interactions. A new kind of industrial espionage.
-
AnthropicAnthropic Said No to the Pentagon, and the Market Said Yes
Anthropic refused to let Claude be used for autonomous weapons. The Pentagon retaliated. The public responded by making Claude the #1 app.