Tech News
← Home  ·  All topics

Mcp

63 GoKawiil briefs on this topic

Blokkily debuts as pattern-first DAW with native CLAP support and LLM assistant

Blokkily is a newly shown experimental digital audio workstation built around a single event model that synchronizes a tracker, piano roll, and step grid as one lane in a multi-track song. It supports CLAP as a required native plugin format alongside VST3 and SoundFont instruments, with sample-accurate parameter automation handled through dedicated plugin adapters, and includes an LLM-based composition assistant accessible via a prompt bar.

New open-source tool REA lets AI agents reverse-engineer apps and binaries

A new toolkit called REA connects coding agents to analysis tools for native binaries, JavaScript/Electron apps, .NET assemblies and websites, letting developers ask an agent to investigate how a feature works and rebuild it. Setup is done via 'npx rea-agents setup' and supports agents including Claude Code, Codex, Cursor, Gemini CLI and Grok Build, with analysis running locally and optional integration with Hopper, Ghidra or IDA for native code.

Edi Life OS launches as self-hosted dashboard with built-in MCP server for AI assistants

A new open-source project called Edi Life OS combines ten workspaces—including habit tracking, Kanban boards, a Pomodoro timer, finance tracking and goal planning—into a single self-hosted dashboard built on PHP and MySQL. The tool includes a built-in MCP server that lets AI assistants like Claude read dashboard data, plan goals, log habits and track expenses on the user's behalf.

Docker Launches Docker Agent for No-Code AI Agent Creation and Deployment

Docker has introduced Docker Agent, a CLI plugin that enables users to build, run, and share AI agents using YAML configurations without coding. The platform supports multi-agent teams, a broad tool ecosystem, and compatibility with various AI providers, all deployable via Docker or OCI registries.

Pi 1.0 adds MCP tool support through new 'Codemode' feature

The developer behind Pi has released version 1.0 with a feature called Codemode, which brings Model Context Protocol (MCP) support into the harness. The post explains how tool-calling works for LLMs, why bash/CLI composition has been favored, and where that approach breaks down for things like image reading or spawning sub-agents.

Security Flaw in MCP Protocol Poses Risks for AI Agent Networks

Researcher Syed Anas Mohiuddin demonstrated that the Model Context Protocol (MCP), used for internal communication among AI agents, has vulnerabilities that can be exploited to spread malicious instructions. Tests across organizations like Google, JP Morgan Chase, and government agencies revealed that trust gaps and lax guardrails enable prompt injection attacks to compromise internal systems. These exploits can lead to unauthorized network requests and data breaches, highlighting a significant security concern for AI-driven environments.

Team's month-long GLM 5.3 Flash coding trial derailed by costs, outages

A development team set out to run an entire month of coding work on the GLM 5.3 Flash model, keeping within a $68 budget for the first two weeks. The second half of the month saw usage spike to 1 billion tokens across other models, driven by an expensive vibe-coded Wagtail MCP server prototype that alone cost $150 and 5kWh, plus infrastructure capacity issues that forced switches to DeepSeek V4.1 Flash and Qwen 3.8 Flash.

New open-source audio editor Audionaut adds MCP support for AI agents

Audionaut is a free, open-source multitrack audio editor built with C++ and JUCE that runs on Windows, macOS, and Linux. It supports MCP (Model Context Protocol), allowing AI agents like Claude to directly edit audio sessions, with each change recorded as a single undo step.

New Pi extension lets any model use Codex's connected apps

WileAI released a Pi extension, @wileai/pi-codex-connector, that lets any Pi-compatible model access connectors already linked to a user's ChatGPT/Codex account, such as GitHub, Gmail, Google Calendar, Slack, Linear and Figma. The extension launches a Codex app-server session with no model attached and exposes just three tools—codex_connectors, codex_connector_schema and codex_connector_call—so a Pi model can discover and invoke Codex connector tools without loading all their schemas directly. Write actions can be set to ask for confirmation, run automatically, or be blocked entirely via the PI_CODEX_CONNECTORS_WRITES setting.

Figma limits remote MCP server access to approved clients, excludes Pi

A Figma representative confirmed that the company's remote MCP server only works with clients on an approved allowlist, and the Pi client is not currently included. Figma pointed to its public MCP catalog listing supported clients and directed developers to a form to request that Pi be added in the future.

Nanoleaf launches MCP server for ChatGPT and Claude control of smart lights

Nanoleaf has released a Model Context Protocol server enabling its smart lights to be controlled through AI assistants like ChatGPT and Claude using natural language commands. The system can also let lights react to real-time data such as weather. Developer tools are available now, with pre-built integrations for major AI assistants arriving in October.

Reco raises $55M to expand into AI agent security

Reco, previously focused on securing SaaS and AI platforms, has raised $55 million and repositioned its product around a context graph that maps connections between AI agents, apps, people, accounts and permissions. CEO Ofer Klein said the shift responds to companies deploying agents faster than they can track them, citing a Fortune 100 customer where Reco found 21,000 unknown agents and a financial firm where an ex-employee's agent had unauthorized Salesforce access.