AI Agent Article Series
News
Latest updates and technical news related to AI Agents.
Claude -p Just Got Its Own Bill — Good Thing I've Already Started Migrating to Codex (and Why Agent View Isn't What I Needed)
Anthropic announced that starting June 15, `claude -p` and Agent SDK usage no longer count against your subscription — they draw from a separate monthly Agent SDK credit and bill at full API rates once exhausted. For someone like me who treats the filesystem as IPC and uses headless mode to wake teammates up, this isn't a price hike — it's a unit-of-billing change. Same week, Anthropic also shipped Agent View — but it solves the worker-pattern problem, not the specialist-pattern problem I actually have.
AI Traffic Is Testing GitHub, and Even Mitchell Hashimoto Is Leaving When Actions Stalls
Ghostty leaving GitHub is not just a sentimental farewell from a long-time open source maintainer. Read alongside GitHub's recent Actions delays, its own 30X capacity planning language, and the surge in agentic development workflows, it points to a new pressure point: AI automation is starting to stress the PR pipeline developers have treated as background infrastructure.
Eight Cents an Hour: Anthropic Will Run Your Agents for You
Anthropic launches Claude Managed Agents in public beta — define your agent logic, and they handle sandboxing, state management, error recovery, and security. At $0.08/hr with an OS-inspired 'brain-hand decoupling' architecture, plus Notion and Rakuten deploying in under a week, is the Agent-as-a-Service era officially here?
Milla Jovovich's AI Memory Tool: The Perfect Score That Didn't Survive the Week
Actress Milla Jovovich and engineer Ben Sigman built MemPalace, an open-source AI memory system, using Claude. It claimed a 100% perfect score on the LongMemEval benchmark and racked up 5,400 GitHub stars in 24 hours. Days later, developers found the scores were measuring the wrong thing entirely.
Claude Code's Full Source Code Leaked: What 512,000 Lines of Code Reveal
Anthropic's Claude Code CLI accidentally shipped a source map in its npm package, exposing the complete TypeScript codebase — 1,900 files, 512,000+ lines. The leak reveals internal model codenames, an autonomous daemon called KAIROS, a hidden AI pet system BUDDY, and an 'Undercover Mode' for hiding AI identity in open-source contributions.
OpenAI Ships Codex Into Claude Code — When Your Competitor Becomes Your Plugin
OpenAI officially released codex-plugin-cc, letting developers call Codex for code review and task delegation directly inside Claude Code. The first time a major AI company has embedded its tool into a competitor's ecosystem.
Are Slash Commands Dead? Claude Code Quietly Merges Commands into Skills
Claude Code v2.1.3 merged Custom Commands into Skills. Commands in .claude/commands/ are now treated as Skills — bringing auto-loading, context consumption, and controllability debates.