MCP
Technology ● Covered in 39 stories + Follow
MCP (Model Context Protocol) is used in recent coverage as a way to expose tools and data connections to AI agents, including business APIs and agent runtime components. This includes adding ChatGPT-specific MCP extensions, grounding MCP agents in source-aware verification, and using MCP to connect agents to governed or curated data (e.g., GraphQL field-level restrictions, unified permission layers, and audited, tamper-evident governance). MCP has also been adopted in development tooling and infrastructure, such as connectors for building apps in-chat, multi-account tool discovery across environments, and serving unified project context to coding agents.
Updated 2 October 2026
Latest developments
October 2026
Earendil releases Pi 1.0, an open-source AI agent framework, and introduces Pi Durable for long-running, crash-recovering agent applications Open source release
- One MCP server used 18,000 tokens before doing anything. Here’s the workaround.
- iphone-use
- Floani
- Pi 1.0 hardens the minimal agent harness
- OpenAI MCP Extensions (GitHub Repo)
- AI agents speed up development — data access slows them down
September 2026
OpenAI announced new ChatGPT workplace and plugin features at Dev Day, expanding app-like extensions and shared collaboration tools Feature update
AWS released an open-source Pizza Bot desktop app that presents background AI agent outputs and approvals in an email-style inbox interface Open source release
- Microsoft Fabric is where AI agents learn how the business works
- Getting the Source Right, Not Just the Fact: Source-Aware Verification for MCP Agents
- MCP gets AI agents into your APIs. It doesn’t decide what they should see.
- Qiagen grounds drug discovery agents in curated knowledge
- Build a multi-account AI agent with AgentCore Gateway and MCP
- CoreSpeed
- Floot MCP
- mcpgawk
- Palma.ai raises $1.8M to bring governed AI agents to more enterprises
- ContextsBase - Backlog for Coding Agents
- 6-7 loops we use everyday to make PostHog self-driving
- Ruby UTCP
- Clueso MCP
- OpenAI president: “The computer should be there to empower you.” So stop retooling software for AI agents
- Statable Analytics
- Agent Harness vs Agent Framework vs MCP: Which Layer Owns the Loop, State, Tools, Permissions, and Recovery
- AWS Introduces Pizza Bot: An Open Source Inbox for Background AI Agents
- Accordio
- Cadenya
- GitWarren
August 2026
FetchSandbox MCP is shared as an MCP tool for verifying fixes to AI integrations Feature update
Multiple organizations release Model Context Protocol (MCP) server implementations for enterprise and web use cases Open source release
- Cosmic Agent Plugins
- Manage agents, tools and skills at scale with AWS Agent Registry
- Eclatira
- Anthropic Opens a Research Preview of the Model Hardware Standard (MHS): A Shared Specification for AI Agents to Safely Operate Physical Devices
- Lovable CTO: The Future of SaaS Is Apps That Agents Can Use
- Radar makes podcasts searchable — and usable by AI agents
- Govern AI agent tool access with Amazon Bedrock AgentCore Gateway
- What happens to your indexed data when Mistral flips the switch?
- Anthropic defined the standards inside Agent Plugins. So why isn’t it helping govern the format?
- FetchSandbox MCP
- Impractical
- The new Firecrawl MCP
July 2026
Relationships
Products & technology
- ContextsBase integrated with this technology · 1 source
- PostHog integrated with this technology · 1 source
- UTCP integrated with this technology · 1 source
- Cadenya integrated with this technology · 1 source
- Accordio integrated with this technology · 1 source
- Pizza Bot integrated with this technology · 1 source
- WeWeb MCP integrated with this technology · 1 source
- Cosmic integrated with this technology · 1 source
- Clueso deploys this technology · 1 source
- GitWarren integrated with this technology · 1 source
- Anthropic integrated with this technology · 1 source
- Particle integrated with this technology · 1 source