Entity · protocol
MCP
protocolactive
mcp-393fa6aa·67 events·first seen Aliases: MCP
Co-occurring entities
More like this (12)
Guides (1)
Recent events (50)
All 67 events →Figma publishes MCP server guide on GitHub
Figma has published an official guide repository for using the Figma MCP (Model Context Protocol) server, written in Python. The repository has accumulated 1,831 stars, indicating meaningful developer interest. This represents Figma's participation in the growing MCP ecosystem for tool integration with AI agents.
AskChem: Claim-centered retrieval infrastructure for chemistry literature with MCP and agent access
AskChem is a new retrieval infrastructure that reframes chemistry literature search around atomic, provenance-carrying claims rather than document rankings, indexing 2.4M claims from 147K papers. Each claim is grounded with a source DOI and verbatim quote, and the system exposes a faceted taxonomy, evidence graph, and living taxonomy for synthesis. The system provides REST, SDK, and MCP access for AI agents, and benchmarking shows that grounding GPT-5.5 in AskChem yields 100% resolvable DOIs versus 88.3% without retrieval. The work is relevant both as a domain-specific RAG infrastructure paper and as an example of MCP-native scientific tooling for AI agents.
Simon Willison walks through adding a custom MCP server to Claude and ChatGPT
Simon Willison documents a practical walkthrough for configuring a custom MCP (Model Context Protocol) server with both Claude and ChatGPT. The post covers the concrete steps required to integrate a self-hosted MCP server into two major AI assistant platforms. This is a practitioner-level guide relevant to the growing MCP ecosystem and cross-platform tool-use patterns.
AG-UI protocol for agent-user interaction in frontend applications gains traction on GitHub
AG-UI (Agent-User Interaction Protocol) is an open-source TypeScript project aimed at standardizing how AI agents integrate with frontend applications. The repository has accumulated 15,007 stars with 48 added today, indicating sustained community interest. It represents an emerging protocol layer for connecting agent backends to user-facing interfaces, complementary to backend-focused protocols like MCP.
jcodemunch-mcp: MCP server for symbol-level code retrieval via tree-sitter AST
jcodemunch-mcp is an open-source MCP server that uses tree-sitter AST parsing to enable precise, symbol-level GitHub code retrieval, claiming 95%+ token cost reduction for code exploration tasks. The tool is compatible with Claude Code, Cursor, and any MCP client. It has accumulated 2,225 GitHub stars with 313B+ tokens reportedly saved across users.
unifi-mcp: MCP server for UniFi network/security suite
A Python-based MCP server implementation exposing UniFi's suite of applications (Network, Protect, Access, Drive) as tool-callable endpoints. The project enables AI agents to interact with UniFi network infrastructure via the Model Context Protocol. It has 581 GitHub stars with modest daily growth, indicating niche but real adoption.
SurfSense: open-source NotebookLM alternative with live web research via MCP
SurfSense is an open-source Python project positioning itself as a NotebookLM alternative, enabling research across live web sources including Reddit, YouTube, Instagram, TikTok, Google Search, and Maps. It exposes functionality through a platform UI, API, or MCP server. The project has accumulated 15,456 GitHub stars with modest daily momentum (+35 today).
Serena: MCP-based semantic code retrieval and editing toolkit for agents
Serena is an open-source Python toolkit that provides an MCP-compatible interface for semantic code retrieval and editing, positioning itself as an IDE layer for AI coding agents. The project has accumulated 26,793 GitHub stars with 85 added today, indicating sustained community traction. It is relevant to the growing ecosystem of MCP-native tooling for agentic software development workflows.
Claude Code 2.1.217 release: subagent concurrency caps, MCP memory fix, and multi-platform bug fixes
Anthropic released Claude Code version 2.1.217 with a notable new cap on concurrently-running subagents (default 20, configurable via CLAUDE_CODE_MAX_CONCURRENT_SUBAGENTS) and a change preventing nested subagent spawning by default. The release also fixes a memory leak in MCP tool output handling, resolves auto-compact failures for Claude Opus 4.8 on Bedrock, and patches several Windows, corporate proxy/TLS, and session isolation bugs. The subagent concurrency and budget enforcement changes are architecturally significant for agentic workflows, preventing unbounded fan-out.
FastMCP Python library for building MCP servers and clients trending on GitHub
FastMCP, a Python library by PrefectHQ for building Model Context Protocol servers and clients, is trending on GitHub with 26,434 total stars and 77 new stars today. The project positions itself as a high-level, Pythonic interface for MCP development. Its traction signals continued ecosystem growth around the MCP standard for AI tool integration.
SigNoz: OpenTelemetry-native observability platform with MCP integration and AI agent support
SigNoz is an open-source observability platform (30k+ GitHub stars, +343 today) that provides logs, metrics, and traces with APM and distributed tracing capabilities. It has added an MCP server integration and a native AI teammate feature in its cloud offering, positioning it for AI agent observability use cases. The trending activity suggests growing adoption in the AI/ML infrastructure space.
Wigolo: local-first web search and crawl tool for AI coding agents via MCP
Wigolo is a TypeScript open-source tool providing local-first web search, fetch, crawl, and research capabilities for AI coding agents over the Model Context Protocol (MCP). It requires no API keys and incurs no per-query cost, positioning itself as a zero-cost alternative to cloud-based search APIs. The project gained 192 stars in a single day during its public beta, suggesting meaningful community traction.
Data Points: Apple sues OpenAI; Meta Muse Spark 1.1; ChatGPT Work; IBM CodeAlchemy; OpenAI Atlas shutdown
A multi-item digest covers five significant AI developments: Apple sued OpenAI alleging trade secret theft via former employees including hardware chief Tang Tan; Meta released Muse Spark 1.1, a multimodal agentic model with 1M-token context and strong tool-use capabilities; OpenAI launched ChatGPT Work, a cloud-based workplace agent competing with Anthropic's Claude Cowork; IBM released CodeAlchemy, a 500B+ token synthetic code dataset with execution traces showing smaller models trained on it outperform those trained on much larger real-code corpora; and OpenAI shut down its Atlas browser in favor of a Chrome extension and desktop integration. These items collectively reflect intensifying competition across agentic products, synthetic data strategies, and legal disputes between major AI players.
mcp-use: open-source fullstack MCP framework for building MCP apps and servers
mcp-use is a TypeScript framework on GitHub for developing MCP (Model Context Protocol) applications targeting ChatGPT and Claude, as well as MCP servers for AI agents. The project has accumulated over 10,000 stars, indicating meaningful community adoption. It represents a tooling layer in the growing MCP ecosystem for agent-tool integration.
Claude Code 2.1.210 release: bug fixes, security hardening, and agent workflow improvements
Anthropic released Claude Code version 2.1.210, a substantial patch addressing over 30 bugs and adding several improvements to the agentic coding tool. Notable fixes include hardening against indirect prompt injection via subagent-read content, correcting git worktree isolation for subagents, fixing hook callback timeouts that caused unattended sessions to stall, and resolving MCP server teardown during mid-session re-syncs. The release also improves auto-mode permission classification by defaulting to Sonnet 5 for external sessions and adds UX enhancements for the multi-agent dashboard.
Vexa: open-source meeting transcription API with MCP server for AI agents
Vexa is an open-source Python project providing a meeting transcription API for Google Meet, Microsoft Teams, and Zoom, featuring auto-join bots and real-time WebSocket transcripts. It includes an MCP server integration, making it usable as a tool source for AI agents. The project is available for self-hosting or as a hosted SaaS and has accumulated 2,502 GitHub stars with 72 added today.
memtrace: Structural memory layer for AI coding agents using bi-temporal graph and MCP
memtrace-public is an open-source Python library providing structural memory for AI coding agents via a bi-temporal graph architecture. It is MCP-native and operates without LLM calls, targeting integrations with Cursor, Claude Code, Codex, and VS Code. The project is early-stage with 407 stars and modest daily traction.
Claude Code 2.1.206 release: background agent upgrades, gateway login, MCP fixes, and UX improvements
Anthropic released Claude Code version 2.1.206 with a range of new features and bug fixes. Notable additions include background agent auto-upgrade after Claude Code updates, /login support for Anthropic-operated public gateway endpoints, /doctor checks for trimming CLAUDE.md files, and auto-allow for git push to configured remotes in /commit-push-pr. The release also fixes numerous bugs including MCP server timeout handling, OAuth re-authentication loops, model picker pricing display errors, and a multi-minute Bedrock startup hang on restricted networks.
Mistral Studio adds version control and governance for prompts and skills
Mistral has launched version control and system-of-record capabilities for prompts and skills within its Studio platform, targeting enterprise AI governance needs. The feature provides immutable versioning, ownership tracking, audit logs, rollback, and CI/CD integration, allowing non-developer domain experts to iterate on production prompts without engineering bottlenecks. Skills are exposed as MCP servers directly from Studio, connecting governed assets to runtime execution. The release addresses a common enterprise pain point: unmanaged, scattered prompt assets that create compliance and traceability risks.
Meta Superintelligence Labs releases Muse Spark 1.1, a multimodal agentic reasoning model with Meta Model API
Meta Superintelligence Labs has released Muse Spark 1.1, a significant upgrade to Muse Spark featuring a 1-million-token context window, strong agentic and computer-use capabilities, and major coding improvements on complex codebases. The model supports multi-agent orchestration, zero-shot generalization to MCP servers and custom tools, and multimodal reasoning including visual-to-code generation and video understanding. Alongside the model release, Meta is launching a public preview of the Meta Model API, giving developers programmatic access for the first time. Safety evaluations were conducted under Meta's Advanced AI Scaling Framework across frontier risk categories.
Notion releases official MCP server for AI tool integration
Notion has published an official Model Context Protocol (MCP) server implementation in TypeScript, enabling AI agents and tools to interact with Notion workspaces via the MCP standard. The repository has accumulated 4,491 stars on GitHub. This represents a major productivity platform adopting MCP as its AI integration layer.
Google Analytics MCP server on GitHub
A Python-based Model Context Protocol (MCP) server for Google Analytics has appeared on GitHub trending, published under the googleanalytics organization. The repository has accumulated 2,606 stars with modest daily growth (+14). This represents an official or semi-official Google Analytics integration point for AI agents and tools using the MCP standard.
Claude Code 2.1.203 release: major background agent reliability fixes and MCP roots integration
Anthropic released Claude Code version 2.1.203, a substantial patch addressing over 25 bugs primarily affecting background agent sessions, subagent orchestration, and worktree isolation. Key fixes include automatic recovery when daemon session tokens go stale, correct subagent state carry-over when returning to sessions, and PATH/environment variable inheritance issues on Windows. The release also adds MCP roots/list integration for working directories and improves streaming responsiveness.
Microsoft releases skills-for-fabric: MCP systems for Microsoft Fabric via CLI, VSCode, and Claude
Microsoft published an open-source Python repository providing a collection of skills and MCP (Model Context Protocol) servers enabling users of CLI, VSCode, and Claude to operate over Microsoft Fabric. The tooling connects AI assistants to Microsoft's data analytics and engineering platform via the MCP protocol. This is a concrete example of MCP adoption by a major enterprise vendor for production data infrastructure.
Microsoft 'skills' repo: MCP servers and agent grounding primitives for coding agents
Microsoft has published an open-source TypeScript repository called 'skills' that provides MCP servers, custom agents, and an Agents.md specification to ground coding agents within SDKs. The project appears to be infrastructure for standardizing how coding agents discover and invoke capabilities. With 2,656 stars and modest daily growth, it is an early-stage but notable contribution to the agent tooling ecosystem from a major lab.
OpenAI Codex app adds MCP command palette and queued follow-ups
OpenAI's Codex app received a minor update adding MCP (Model Context Protocol) and personality actions to the command palette, along with a change to queue follow-up interactions by default. The update also includes unspecified performance improvements and bug fixes. The MCP integration is the most notable element, continuing the trend of MCP adoption across AI coding tools.
OpenAI Codex app adds MCP shortcuts and review mention support
OpenAI's Codex app received a feature update adding MCP shortcuts in the composer—including install keyword suggestions and an MCP server submenu—alongside @mention and skill mention support in inline review comments. The release also includes improved rendering of MCP tool calls and Mermaid diagram error handling. The update signals continued integration of the Model Context Protocol into OpenAI's developer tooling.
OpenAI Codex adds plugin system with MCP server configuration support
OpenAI has shipped a plugin system for Codex that allows installable bundles packaging skills, app integrations, and MCP server configuration for reusable workflows. Plugins are available across the Codex app, CLI, and IDE extensions, and can be installed per-user or per-repository. The system supports both a curated public plugin directory and local plugin scaffolding via a built-in @plugin-creator skill. The inclusion of MCP server configuration as a first-class plugin component signals continued convergence on MCP as a standard for tool integration in agentic coding environments.
OpenAI adds curated plugins directory to Codex for reusable workflows
OpenAI has introduced a plugins directory in Codex that allows users to discover, install, and share packaged workflows combining skills, app integrations, and MCP server configurations. Plugins are installable bundles designed to make reusable Codex setups portable across projects and teams. Workspace owners retain control over plugin availability through existing app access controls.
OpenAI adds curated plugins directory to Codex
OpenAI has added a plugins directory to Codex, allowing users to discover, install, and use packaged workflows built from apps and skills. Plugins bundle skills, app integrations, and MCP server configurations into reusable units that can be shared across projects or teams. The feature extends Codex's composability and signals continued investment in MCP as a packaging primitive.
OpenAI adds curated plugins directory to Codex with MCP server support
OpenAI has added a plugins system to Codex that allows users to discover, install, and share packaged workflows combining skills, app integrations, and MCP server configurations. Plugins are installable bundles designed to make reusable Codex workflows portable across projects and teams. Enterprise and Edu admins can manage plugin access via workspace controls and RBAC. The integration of MCP server configurations into the plugin format is notable for the agent-tool ecosystem.
OpenAI Codex app update: workspace settings, Windows native support, MCP approval panels, and agent composer features
OpenAI shipped a Codex app update adding workspace settings, native Windows updater with MSIX and system tray support, and MCP custom approval panel handling. The automation composer gains app/file @mentions, subagent diff stats, and artifact cards for file citations. The release also includes a range of bug fixes covering multi-window issues, review panel behavior, and thread search performance.
OpenAI launches ChatGPT Workspace Agents for Enterprise and Business tiers
OpenAI is rolling out Workspace Agents to ChatGPT Business and Enterprise customers, enabling organizations to build, schedule, and share agents for repeatable business workflows. Agents can connect to external apps including Google Drive, Google Calendar, Slack, and SharePoint, and support custom MCP servers. The feature includes template-based creation, version history, analytics, and admin controls, with Slack integration for running agents in connected channels.
OpenAI launches ChatGPT Workspace Agents for Enterprise and Business tiers
OpenAI is rolling out Workspace Agents to ChatGPT Business and Enterprise customers, enabling organizations to build, schedule, and share agents for repeatable business workflows. Agents can connect to tools including Google Drive, Google Calendar, Slack, and SharePoint, and support custom MCP servers. The feature ships off-by-default with admin controls, and is excluded from EKM-enabled Enterprise workspaces at launch. This represents OpenAI's direct push into enterprise workflow automation, competing with established RPA and no-code automation platforms.
OpenAI Codex app update: realtime voice context and MCP panel stability fixes
OpenAI released a minor update to the Codex app adding a tooltip clarifying that realtime delegation messages use surrounding voice conversation as context. The update also fixes search navigation in long review files, prevents embedded MCP app panels from losing state during fullscreen changes, and resolves several desktop regressions including tray crashes and macOS hotkey issues.
OpenAI releases GPT-5.5 and GPT-5.5 Pro to the API with 1M token context and built-in agentic tools
OpenAI released GPT-5.5 and GPT-5.5 Pro to the Chat Completions and Responses API, positioning them as frontier models for complex professional work and compute-intensive tasks respectively. GPT-5.5 supports a 1M token context window, image input, structured outputs, function calling, built-in computer use, hosted shell, MCP, web search, and Skills. Notable behavioral changes include reasoning effort defaulting to medium and extended-only prompt caching support.
ChatGPT Workspace Agents now available for Enterprise workspaces with EKM
OpenAI has extended ChatGPT Workspace Agents to eligible Enterprise workspaces using Enterprise Key Management (EKM). The feature allows organizations to build, schedule, and share agents for repeatable business workflows across connected apps including Slack, with support for custom MCP servers, skills, files, and version history. Agents are off by default and require admin enablement, signaling a cautious enterprise rollout posture.
OpenAI releases Secure MCP Tunnel for enterprise customers
OpenAI has released Secure MCP Tunnel, an enterprise feature enabling supported products—ChatGPT web, Codex, Responses API, and AgentKit—to connect to private or on-premises MCP servers via a customer-hosted tunnel client without exposing those servers to the public internet. The feature addresses a key enterprise adoption barrier by allowing MCP-based tool integrations to remain within private network boundaries. This is a meaningful infrastructure addition for organizations deploying AI agents against internal systems.
OpenAI Codex app adds Appshots, graduates Goal mode, and enables remote computer use
OpenAI's Codex app for macOS received a significant feature update including Appshots (a hotkey to send the frontmost window's screenshot and text directly to Codex), graduation of Goal mode from experimental to stable across app/IDE/CLI, and remote computer use that allows Codex to operate desktop apps while the Mac is locked. The release also adds plugin sharing via marketplace for ChatGPT Business, browser-use improvements including faster image asset extraction and a read-only JS sandbox for structured data, and various reliability fixes. These features collectively push Codex toward longer-horizon autonomous agentic operation with expanded environmental access.
ChatGPT adds enterprise app templates for GitHub Enterprise, Snowflake, and Databricks
OpenAI has released workspace app templates for ChatGPT Business plan admins, enabling guided setup of integrations with GitHub Enterprise, Snowflake, and Databricks. The templates handle OAuth credentials, callback URLs, webhook details, and managed MCP server URLs, then allow admins to publish and manage apps with role-based access controls. This extends ChatGPT's enterprise connectivity to major data and developer platforms via a standardized MCP-backed integration pattern.
ChatGPT adds enterprise app templates for GitHub Enterprise, Snowflake, and Databricks
OpenAI has released app templates in ChatGPT for Enterprise and Edu plan workspace admins, enabling guided setup of workspace-specific integrations with GitHub Enterprise, Snowflake, and Databricks. The templates handle OAuth credentials, callback URLs, webhook details, and managed MCP server URLs, then allow admins to publish and manage apps with role access and action controls. This extends ChatGPT's enterprise connectivity to major data and developer platforms via a structured provisioning flow.
Claude Code 2.1.178: Simplified agent teams, parameter-based permission rules, and subagent safety improvements
Claude Code version 2.1.178 overhauls the experimental agent teams feature by removing explicit TeamCreate/TeamDelete tools in favor of an implicit single-team model where subagents are spawned directly via the Agent tool's name parameter. A new Tool(param:value) permission syntax allows matching tool invocations by input parameter values (e.g., blocking Opus subagents specifically). A notable safety fix closes a gap where subagents could request blocked actions without classifier review before launch. The release also includes numerous bug fixes for subagent transcript visibility, auth token handling, compaction fallback, and nested .claude/ directory resolution.
Claude Code 2.1.181 release: /config syntax, Apple Events sandbox, subagent improvements, and ~40 bug fixes
Anthropic released Claude Code version 2.1.181 with several new features including a /config key=value prompt syntax for setting any configuration option, an opt-in Apple Events sandbox permission on macOS, and a new CLAUDE_CLIENT_PRESENCE_FILE environment variable for suppressing mobile push notifications. The release also includes improvements to subagent panel UX (auto-hide, depth limits, elapsed time fixes), streaming of long paragraphs, and auto-retry on mid-thinking connection drops. Approximately 40 bug fixes address issues ranging from 0-byte file writes on network drives, macOS TUI freezes, startup regressions, and credential refresh loops.
Claude Code 2.1.183 adds destructive-command safety blocks and multiple bug fixes
Claude Code version 2.1.183 introduces auto-mode safety guardrails that block destructive git and infrastructure commands (git reset --hard, terraform destroy, etc.) unless explicitly requested, and prevents amending commits not made by the agent in the current session. The release also adds a deprecated-model warning on stderr, a new attribution.sessionUrl config option, and fixes a range of bugs including empty WebSearch results in subagents, fullscreen TUI corruption on Windows Terminal, MCP auth-stub tool exposure in headless mode, and scheduled task triggers incorrectly acting as keyboard input.
Claude Code 2.1.186 adds MCP CLI auth, agent-team improvements, and numerous fixes
Claude Code version 2.1.186 ships new CLI commands for authenticating MCP servers without the interactive menu (`claude mcp login/logout`), including SSH-compatible stdin redirect support. Agent-team behavior is improved: background subagents now surface permission prompts in the main session rather than auto-denying, teammate spawns inherit the leader's effort level, and subagents looping on schema validation failures now abort after 5 attempts. The release also fixes over 20 bugs spanning streaming reliability, subagent transcript handling, Chrome tab-group isolation, and session cost display for Enterprise/Team subscribers.
Claude Code 2.1.187 release: sandbox credentials, org model restrictions, MCP timeout fixes
Anthropic released Claude Code version 2.1.187 with a mix of security, reliability, and UX improvements. Notable additions include a sandbox.credentials setting to block credential file access in sandboxed commands, org-level model restriction enforcement in the model picker, and mouse click support in fullscreen menus. Key bug fixes address remote MCP tool calls hanging indefinitely (now abort after 5 minutes), subagent depth tracking on resume, leaked agent worktree registrations, and CJK text mojibake in certain terminals.
Claude Code 2.1.191 release: /rewind support, MCP reliability improvements, 37% CPU reduction
Anthropic shipped Claude Code version 2.1.191 with a notable set of bug fixes and performance improvements. Key additions include /rewind support for resuming conversations before a /clear, roughly 37% CPU reduction during streaming via coalesced text updates, and improved MCP server reliability with retry logic for transient network errors. Several agent-related bugs were fixed, including background agents resurrecting after being stopped and slash commands being incorrectly forwarded to background sessions.
Claude Code 2.1.193 adds shell classifier routing, OTel response logging, and background agent fixes
Claude Code version 2.1.193 ships several notable additions: a new autoMode.classifyAllShell setting to route all shell commands through the auto-mode classifier, an OpenTelemetry log event for assistant responses (redacted by default, with opt-in via environment variable), and live file path autocomplete in bash mode. The release also fixes multiple background agent bugs including phantom subagent spawning, spurious cancellation on backgrounding, and agent panel display issues, while improving MCP authentication reconnection on 401/403 errors.
Claude Code 2.1.196 release: org-default models, MCP security fix, background agent reliability improvements
Anthropic released Claude Code version 2.1.196 with a notable security fix preventing untrusted workspace .mcp.json servers from auto-spawning, and added support for organization-level default model configuration via the admin console. The release includes significant background agent reliability improvements — long-running workflows now survive process restarts on all platforms including Windows — plus a streaming idle watchdog enabled by default that aborts and retries stalled response streams after 5 minutes. Numerous bug fixes address agent session state management, MCP OAuth scope handling, PowerShell compatibility, and a /code-review token usage reduction of ~25%.
Anthropic launches Claude Science, an AI workbench for scientific research
Anthropic released Claude Science in beta, an AI-powered research environment integrating over 60 curated scientific tools and databases for genomics, proteomics, structural biology, and cheminformatics. The platform features a coordinating agent with specialist sub-agents, a reviewer agent for citation and calculation checking, reproducible auditable artifacts, and flexible compute management across local machines, HPC clusters, and on-demand GPUs. It integrates with NVIDIA's BioNeMo Agent Toolkit, connecting to models like Evo 2, Boltz-2, and OpenFold3. Available to Claude Pro, Max, Team, and Enterprise users, this represents Anthropic's most significant expansion into scientific AI tooling.