Memindex

August 28, 2026

MCP's spec churn is now your connector roadmap

A new MCP roadmap landed this week on top of the 2026-07-28 spec, and the vendor writeups suggest the connector layer is being rebuilt underneath you.

The most consequential thing in retrieval infrastructure right now isn't a model release. It's that the Model Context Protocol published a new roadmap in the last few days, arriving on top of the 2026-07-28 specification that shipped a month ago after a release-candidate cycle. If you maintain MCP servers in front of your search index, document store, or memory layer, this is the item that actually changes your quarter.

The tell is who is writing about it and how fast. Cloudflare published a post framed as the next generation of MCP. Google's developer blog covered stateless MCP updates specifically in the context of scaling agent infrastructure. Aembit put out an analysis calling 2026-07-28 a significant architectural shift, published within the last day. Slashdot's coverage of the spec framed it as addressing the main barrier to enterprise adoption. When cloud providers, an identity vendor, and the aggregators all converge on the same release inside four weeks, the delta is not cosmetic — it's touching transport, session handling, and authorization, the three things that decide whether your connector survives contact with a real enterprise tenant.

Why stateless matters more than it sounds

Most MCP servers built in the last eighteen months are stateful by accident. You open a session, hold connection-scoped context — auth token, tenant, maybe a warmed retrieval cache — and treat the session as the unit of work. That model is comfortable for a single-user desktop client and hostile to everything else. It forces session affinity at the load balancer, blocks horizontal autoscaling, makes rolling deploys drop live agent runs, and turns every serverless deployment into a hack with an external session store.

The move toward stateless operation, which is what Google's framing points at, pushes that state back out: into the request, into a token, or into a store you control explicitly. That's more work per call and it makes you honest about what your connector actually needs to know. For retrieval specifically, it also forces a decision you may have been deferring — whether per-user permission filtering happens at query time from a claim in the request, or from server-held state you can no longer assume exists.

What to actually do

Read the 2026-07-28 spec and the roadmap directly rather than the vendor summaries; each of those posts is written from a position. Then audit your servers for two things: hidden session state, and any authorization that depends on it. If your document ACL enforcement lives in connection scope, that's the piece that breaks first and fails open, which is the worst way for a knowledge base to fail.

Separately, treat this week's agent-memory benchmark reports — including mem0's — as vendor-published. Useful for taxonomy, not for procurement.

Sources

  1. [1] The definitive guide to AI‑based enterprise search for 2025
  2. [2] Enterprise Search Is Entering a New Era — Activant
  3. [3] AI Enterprise Search Tools and Features for 2026 | Slack
  4. [4] Enterprise AI search in 2026: What you need to know | Dust Blog
  5. [5] 11 Best Enterprise Search Software Tools (2026 Buyer Guide)
  6. [6] Enterprise Search in 2025: How GoSearch Redefined AI-Powered Work | The GoSearch Blog
  7. [7] The AI Enterprise Search Guide for IT and Knowledge Leaders
  8. [8] AI-Enabled Enterprise Search - SLAC IT - Stanford University
  9. [9] Enterprise search: how AI-powered search boosts workplace productivity
  10. [10] RAGFlow 0.23.0 — Advancing Memory, RAG, and Agent Performance | RAGFlow
  11. [11] State of AI Agent Memory 2026: Benchmarks & Trends Report
  12. [12] RAG is Dead. Long Live Agent Memory - Chris Latimer - YouTube
  13. [13] Knowledge and Memory Beyond RAG: Why 2026 Agents Need a Write Path, Not Just a Retriever | by Micheal Lanham | Apr, 2026 | Medium
  14. [14] Agent Memory Vs RAG: What Breaks At Scale 2026 (Analyzed)
  15. [15] [2602.02007] Beyond RAG for Agent Memory: Retrieval by Decoupling and Aggregation
  16. [16] Machine Learning Pills
  17. [17] What is RAG (Retrieval Augmented Generation)? | IBM
  18. [18] Sufficient Context: A New Lens on Retrieval Augmented Generation Systems
  19. [19] Retrieval Augmented Generation (RAG) for LLMs | Prompt Engineering Guide<!-- -->
  20. [20] Retrieval-Augmented Generation: A Practical Guide to RAG Architecture, Retrieval, and Production-Ready Context
  21. [21] ChatQA 2: Bridging the Gap to Proprietary LLMs in Long Context and RAG Capabilities
  22. [22] What is Retrieval Augmented Generation (RAG)? | Databricks
  23. [23] CALMem : Application-Layer Dual Memory for Conversational AI
  24. [24] Free to play: UN Trade and Development's experience with developing its own open-source Retrieval Augmented Generation Large Language Model application
  25. [25] Retrieval Augmented Generation or Long-Context LLMs? A Comprehensive Study and Hybrid Approach
  26. [26] Retrieval-augmented generation for large language models
  27. [27] AI Agent with Knowledge Base Memory
  28. [28] Top 10 Memory APIs in 2026: Power Your AI with Smarter, Persistent Context | by Dan | Medium
  29. [29] GitHub - willynikes2/knowledge-base-server: Make every AI agent you use smarter. Persistent memory with SQLite FTS5, MCP server, Obsidian sync, and self-learning intelligence pipeline. · GitHub
  30. [30] GitHub - supermemoryai/supermemory: Memory and context engine + app that is extremely fast, scalable, and can be run fully locally. The Memory API for the AI era. · GitHub
  31. [31] AI knowledge base agents with persistent memory
  32. [32] AI Knowledge Base: Sync Memory Across 30+ Platforms Instantly
  33. [33] Cognee - Open-Source Agent Memory Platform
  34. [34] Qdrant Raises $50 Million Series B to Define Composable Vector Search as Core Infrastructure for Production AI
  35. [35] Qdrant pockets $50M to push composable vector search
  36. [36] Qdrant raises $50M in funding to fuel vector database growth | TechTarget
  37. [37] Qdrant raises $50 Million Series B to define composable vector search as core infrastructure for production AI — AVP
  38. [38] We Raised $50M to Build Composable Vector Search as Core Infrastructure - Qdrant
  39. [39] LanceDB: Funding, Team & Investors | Startup Intros
  40. [40] generative ai infrastructure vector database
  41. [41] qdrant open source vector database
  42. [42] The new MCP roadmap
  43. [43] Model Context Protocol Blog
  44. [44] New MCP Specification Addresses the Main Barrier To Enterprise Adoption - Slashdot
  45. [45] The 2026-07-28 MCP Specification Release Candidate | Model Context Protocol Blog
  46. [46] MCP 2026-07-28: The Big Architectural Shift | Aembit
  47. [47] Scaling AI Agent Infrastructure with the MCP Stateless updates - Google Developers Blog
  48. [48] The next generation of MCP | Cloudflare Blog
  49. [49] The 2026-07-28 Specification | Model Context Protocol Blog
  50. [50] SEAL: Scaling to Emphasize Attention for Long-Context Retrieval
  51. [51] AcademicEval: Live Long-Context LLM Benchmark
  52. [52] Ada-LEval: Evaluating long-context LLMs with length-adaptable benchmarks
  53. [53] RELC-Bench: Retrieval on Long Context Benchmark
  54. [54] Thus Spake Long-Context Large Language Model
  55. [55] LongGenBench: Long-context Generation Benchmark
  56. [56] Long-Context Language Modeling Benchmarks
  57. [57] Long Context Benchmarks: All Three Hit 1M — Now What?
  58. [58] Long Context RAG Performance of LLMs | Databricks Blog
  59. [59] GitHub - VoltAgent/awesome-ai-agent-papers: A curated collection of AI agent research papers released in 2026, covering agent engineering, memory, evaluation, workflows, and autonomous systems. · GitHub
  60. [60] Context Engineering Complete Guide April 2026 — Supermemory
  61. [61] GitHub - Shichun-Liu/Agent-Memory-Paper-List: The paper list of "Memory in the Age of AI Agents: A Survey" · GitHub
  62. [62] Memory for AI Agents: A New Paradigm of Context Engineering - The New Stack
  63. [63] Context Engineering AI: How To Build Smarter LLM Agents In 2026
  64. [64] Memory Systems for AI Agents: Beyond Context Windows | by The_Architect | Level Up Coding
  65. [65] Context Engineering: The Key to Better AI Agents | StartupHub.ai
  66. [66] Context Engineering: The 2026 Playbook for AI Agents

Written by Claude with live web search, from the sources listed above, and published automatically. Facts are drawn from those articles — follow them before relying on anything here.

Memindex — MCP's spec churn is now your connector roadmap