CLAUDE CODE

Anthropic confirms that every Claude model released after August 2, 2026 embeds a machine-readable watermark in generated text and attaches C2PA provenance metadata to generated files, across the API, Claude Code, Cowork, and Tag. Detection tooling for third parties is coming, but details are not published yet.

Claude Code v2.1.224 lets one running session message another over a first-party channel - plain text, permission-aware, with approval dialogs when bypass-mode sessions talk to each other. Here is what ships, how delivery and inbound controls work, and where the feature stops.

Prime Intellect open-sourced Prime Agent on August 5, 2026. It gives the model exactly one tool - a persistent IPython kernel - and lets the harness rewrite its own prompts, skills, memory, and sub-agents mid-run. Here is how it works, what the benchmarks actually show, a full provider and model guide, and an honest comparison to Claude Code, Codex, OpenCode, OpenClaw, Hermes, and Pi.

A new arXiv benchmark ran 6,560 sandboxed runs across Claude Code, Codex, OpenClaw, and Hermes with five LLMs. 68% of runs triggered unsafe signals, and 66% of all runs were unsafe yet still passed completion checks. Task completion does not prove an agent ran safely.

Google's Antigravity CLI replaced Gemini CLI on June 18, 2026. Here is how it compares to Claude Code and Codex on architecture, pricing, multi-agent workflows, and daily coding experience.

A controlled ablation across Claude Code and Codex, 17 real tasks, and 288 evaluated runs finds context-injection strategy does not measurably change correctness (bounded to under 10-15pp). The failures are implementation skill, not missing repository knowledge.

A July 2026 Microsoft study of Claude Code and GitHub Copilot CLI found roughly 24% more merged pull requests among adopters, but the interesting lesson is rollout design, not magic productivity.

Agent-Manager wraps tmux into a Go TUI that groups AI coding agents by project, shows live status for each, and lets you answer blocked agents or review their changes without attaching to their terminal.

Agent-Manager, Pane, and Golutra let you run multiple CLI coding agents in parallel. Here is the comparison of architectures, agent support, and which fits your workflow.

Effort levels and model choice both cost more for more capability, but they are not interchangeable. Here is when to move the effort dial and when to switch models instead.

How Claude Code, Cursor, Codex, GitHub Copilot, Aider, and Windsurf handle permissions, sandboxing, credential protection, and prompt injection. A structured comparison for engineering teams evaluating agent security.

Anthropic cut 80% of Claude Code's system prompt for Opus 5 and Fable 5 with zero regression on coding evals. The post landed on HN with 197 points and 133 comments. Here is what the article says, what HN thinks, and what it means for your agent harness.

The late-July Codex and Claude Code updates point in the same direction: coding agents are competing on approval modes, resumable work, MCP auth, artifacts, and review surfaces as much as raw model quality.

Anthropic removed over 80% of Claude Code's system prompt for Claude 5 models. Here is how the rules changed and what it means for your CLAUDE.md files, skills, and system prompts.

Open Design lets you point Cursor or Claude Code at any live website and pull out a brand-ready DESIGN.md with colors, typography, spacing, and voice -- no manual extraction, no guesswork, all Apache-2.0.

Agents that critique their own output, learn from mistakes, and get better over time - the three patterns that actually ship, from simple reflection loops to tree search and meta agents.

A July 2026 paper shows how hallucinated repository and skill names can become promptware delivery paths. The practical fix is boring: search before fetch, verify names, and sandbox every install.

Prompt injection, sandbox escapes, and hallucinated dependencies are now documented, patched, CVE-numbered realities. Here is the threat model for agent-written code and the defenses worth adopting this week, ranked by effort.

A step-by-step guide to configuring an isolated Mac that Claude Code can fully control remotely - from SSH and Dispatch to phone-based control with Remote Control.

How a 60-second auto-continue timer shipped to Claude Code without documentation, what it reveals about agent safety assumptions, and how to disable it.