> ## Content Index > Fetch the complete content index at: https://ai-news.ghost.io/llms.txt > Use this file to discover other available public pages before exploring further. # MCP's 2026-07-28 spec goes stateless β€” what breaks now and what's on a 12-month clock - URL: https://ai-news.ghost.io/mcp-goes-stateless/ - Published: 2026-07-29T18:18:48.000Z - Updated: 2026-08-07T06:08:05.000Z - Description: MCP goes stateless; Codex's 5-hour cap returns today; Copilot drops two Gemini models Thursday; Claude finds new crypto attacks on its own. - Author: The New Way The Model Context Protocol's biggest rewrite since launch went final this week β€” the stateless core, with several features removed outright and three more on a twelve-month deprecation clock. OpenAI reset Codex and ChatGPT Work usage limits again and says the temporarily-lifted five-hour cap comes back today. Anthropic says its research model found new attacks against a post-quantum signature scheme and a reduced-round version of AES, entirely on its own. And GitHub's Copilot drops two Gemini models from every surface in two days. The daily pulse of AI coding tools β€” what shipped, what matters, what's next. πŸ“Œ ****In this issue** β€’ [MCP goes stateless β€” some features removed now, more on a 12-month clock](https://ai-news.ghost.io/mcp-goes-stateless/#mcps-2026-07-28-spec-goes-stateless-%E2%80%94-what-breaks-now-and-whats-on-a-12-month-clock) β€’ [Codex limits reset β€” the 5-hour cap returns today](https://ai-news.ghost.io/mcp-goes-stateless/#openai-resets-codex-usage-limits-again-%E2%80%94-tibo-says-the-5-hour-cap-returns-today) β€’ [Copilot drops two Gemini models Thursday](https://ai-news.ghost.io/mcp-goes-stateless/#github-drops-gemini-25-pro-and-gemini-3-flash-from-every-copilot-surface-in-two-days) β€’ [Claude found new crypto attacks on its own, Anthropic says](https://ai-news.ghost.io/mcp-goes-stateless/#anthropic-says-claude-found-new-attacks-on-a-post-quantum-cipher-and-aes-without-human-help) --- ## The protocol every agent depends on finishes its rewrite ### MCP's 2026-07-28 spec goes stateless β€” what breaks now and what's on a 12-month clock The Model Context Protocol β€” the plug standard AI coding agents use to reach external tools β€” dropped its `initialize` handshake and session header; per the spec's own blog, "any request can land on any instance behind a plain round-robin load balancer." No sticky sessions, no shared storage β€” the setup self-hosted MCP servers needed before. **What breaks on upgrade:** `ping`, `logging/setLevel`, and SSE stream-resume are removed, and a new `server/discover` RPC is mandatory. **What's on a clock:** Roots, Sampling, and Logging are deprecated β€” "they'll keep working for at least twelve months" β€” along with Dynamic Client Registration, superseded by Client ID Metadata Documents (a client publishes its own registration data at a URL instead of registering with each server). The removals break your build today; the deprecations put a clock on code you already shipped. [The 2026-07-28 SpecificationThe 2026-07-28 Model Context Protocol specification is out, bringing a stateless protocol core, Multi Round-Trip Requests, header-based routing, cacheable list results, authorization hardening, a formal extensions framework, and updated Tier 1 SDKs.![](https://storage.ghost.io/c/9c/1f/9c1ff71c-2859-44c2-a962-1337982a6484/content/images/icon/favicon-0e7e1194-0475-4370-adb2-7bf3f12d898d.svg)Model Context Protocol BlogDavid Soria Parra (Lead Maintainer) and Den Delimarsky (Lead Maintainer)![](https://storage.ghost.io/c/9c/1f/9c1ff71c-2859-44c2-a962-1337982a6484/content/images/thumbnail/og-image-01e02a93-a644-4fb4-acb6-17a684fb6b4a.png)](https://blog.modelcontextprotocol.io/posts/2026-07-28/?ref=ai-news.ghost.io) Changelog: [modelcontextprotocol.io/specification/2026-07-28/changelog](https://modelcontextprotocol.io/specification/2026-07-28/changelog?ref=ai-news.ghost.io) Β· Discussion: [Hacker News (124 points)](https://news.ycombinator.com/item?id=49088058&ref=ai-news.ghost.io). One oddity as we publish: the spec's own versioning page still lists 2025-11-25 as current, even though the new pages are live. --- ## Two usage-limit stories, two vendors ### OpenAI resets Codex usage limits again β€” Tibo says the 5-hour cap returns today Tibo Sottiaux, OpenAI's Codex lead, says efficiency fixes should make GPT-5.6 Sol usage "last around 18% longer" and that OpenAI has "not reduced usage on any subscription plans" β€” the drain was Sol working harder (more tool calls, code mode, parallel work), and it hit power users on hard tasks far more than the median user. **If you've been running long Codex sessions assuming no 5-hour wall since the July 12 lift, it's back today.** > Hello people of Sol! I've reset usage limits for all ChatGPT Work and Codex users. Together with that, a quick update on GPT-5.6 Sol usage limits. > > Over the past few weeks, many of you have told us that Sol was using your Codex limits faster than expected. To be clear, we have… > > β€” Tibo (@thsottiaux) [July 29, 2026](https://x.com/thsottiaux/status/2082317452755751098?ref%5Fsrc=twsrc%5Etfw&ref=ai-news.ghost.io) --- ### GitHub drops Gemini 2.5 Pro and Gemini 3 Flash from every Copilot surface in two days Announced July 2, effective Thursday July 31: Copilot Chat, inline edits, ask and agent modes, and code completion all lose both models. GitHub's advice: switch to Gemini 3.1 Pro or Gemini 3.5 Flash β€” Enterprise admins may need to enable the replacements in their Copilot model policy first. **A config default that moves: anything pinned to the old model names breaks silently on Thursday.** [Upcoming deprecation of Gemini 2.5 Pro and Gemini 3 Flash - GitHub ChangelogWe will deprecate Gemini 2.5 Pro and Gemini 3 Flash across all GitHub Copilot experiences (including Copilot Chat, inline edits, ask and agent modes, and code completions) on July 31st,…![](https://storage.ghost.io/c/9c/1f/9c1ff71c-2859-44c2-a962-1337982a6484/content/images/icon/cropped-github-favicon-512-8993014a-9ef1-4682-8098-3c0059c10dab.png)The GitHub BlogAllison![](https://storage.ghost.io/c/9c/1f/9c1ff71c-2859-44c2-a962-1337982a6484/content/images/thumbnail/Gemini-models-deprecation-social-580cb395-38cd-4abf-8aec-fe8a97a90e96.jpg)](https://github.blog/changelog/2026-07-02-upcoming-deprecation-of-gemini-2-5-pro-and-gemini-3-flash/?ref=ai-news.ghost.io) --- ## Anthropic's research team, not its product team ### Anthropic says Claude found new attacks on a post-quantum cipher and AES, without human help Using an internal model called Claude Mythos Preview, Anthropic reports a practical key-recovery attack on HAWK-256 β€” a post-quantum signature scheme under NIST review, deployed nowhere β€” cutting the expected attack cost from 2⁢⁴ to 2³⁸ operations, plus a 200–800x speedup on an existing attack against a reduced 7-round version of AES-128 (the full 10-round cipher is unaffected). Each finding cost roughly $100,000 in API spend; HAWK's authors were notified in June, with disclosure coordinated to the public NIST mailing list. **Not a coding-tool release β€” but when someone tells you "Claude broke a cipher," this is what actually happened, and what didn't: nothing deployed is affected.** [Discovering cryptographic weaknesses with ClaudeAnthropic researchers find weaknesses in cryptographic algorithms with Claude Mythos Preview![](https://storage.ghost.io/c/9c/1f/9c1ff71c-2859-44c2-a962-1337982a6484/content/images/icon/favicon-32x32-1596fc8e-17ab-41aa-8a3d-e947aa46825a.png)![](https://storage.ghost.io/c/9c/1f/9c1ff71c-2859-44c2-a962-1337982a6484/content/images/thumbnail/opengraph-illustration-9f2fbb0f-df6f-4229-a9ec-7380f1eb02fc)](https://www.anthropic.com/research/discovering-cryptographic-weaknesses?ref=ai-news.ghost.io) --- That's Wednesday. If the MCP migration breaks something real in your stack, reply β€” what breaks in practice beats what the changelog says should. And if someone forwarded you this, [subscribe here](https://ai-news.ghost.io/#/portal/signup). ## Also worth your time [Show HN: Formally verified 3D CSG β€” trust 93 lines of spec, not 1,000 lines of AI code](https://news.ycombinator.com/item?id=49083239&ref=ai-news.ghost.io) β€” the AI wrote \~1,000 lines of implementation plus 60,000 lines of Lean 4 proofs; the proof checker verifies all of it against a 93-line human-readable spec, so nobody reads either pile to trust the result. (111 points; [repo](https://github.com/schildep/verified-3d-mesh-intersection?ref=ai-news.ghost.io)) [OpenAI is retiring Atlas, its browser, on August 9](https://help.openai.com/en/articles/20001371-evolving-atlas-into-chatgpt-for-browser-based-agentic-work?ref=ai-news.ghost.io) β€” browser-agent capability moves into ChatGPT and Codex directly. Bookmarks, tabs, and history don't migrate automatically; export before the cutoff. --- *The New Way is human-curated β€” a person picks every story. The summaries are written with AI (Claude) and reviewed before we hit send.*