Tag
A benchmark study comparing MCP and filesystem access for AI agents in 20 production scenarios found that filesystem access reduces LLM costs by 27% and latency by 32% while improving answer quality.
The author discusses how AI agents' reasoning degrades with too many tools, noting a cap of 20 tools by Genie Code, and invites community discussion on solutions.
This is a complete guide to using Codex and MCP for AI workflows, covering practical applications, personal workflows, skill creation, AI Agent team building, Agent development, and industry insights and business models.
A hosted MCP server provides a unified API for agent-driven social media management, enabling automated content workflows with human approval gates and platform-specific handling.
ElevenLabs MCP in Claude allows users to create and manage voice agents within chat systems, integrating ElevenLabs' voice technology with Claude's AI capabilities.
Blender Agent Bridge is an open-source MCP bridge designed to integrate AI workflows into Blender, enhancing 3D modeling capabilities with AI tools.
Hermes Agent Desktop has upgraded its capabilities tab, enabling users to scope skills, tools, and MCPs to profiles or bots and install skills directly through a skills browser.
The tweet discusses the lack of safety mechanisms for irreversible tool calls in AI agents and introduces OrcaRouter's new agent firewall feature that scores tool/MCP calls before execution, while emphasizing the need for independent evaluation.
Kody v2026.08.15 is released, enabling package apps, services, and subscriptions to call attached remotes in this AI assistant platform built on Cloudflare Workers.
FetchSandbox MCP is a development tool that validates AI integration fixes, helping developers ensure their changes work correctly in AI systems.
Mole is a deep research agent for the terminal that uses AI to decompose questions, search sources, and provide answers with verified citations while enforcing budgets and ensuring local data privacy.
A video explains the inner workings of AI agents using the open-source Obsidian plugin Vault Operator as a real example, and announces its 3.5.0 release with a redesigned sidebar and skill registry.
A discussion proposing that AI agents working with production data should use data branches — letting the agent write to a cloned database and then reviewing the final diff instead of approving each action individually.
Agent Plugins is a new open, vendor-neutral standard for packaging Agent Skills and their MCP dependencies into one portable folder, with Google joining as a core maintainer. The article details how the standard simplifies distribution, its manifest requirements, and its compatibility across clients.
MCP-Memory is an open-source MCP server that gives AI agents persistent memory using Google's Open Knowledge Format and SQLite FTS5 for fast retrieval and search.
Introduces the open-source project Loop Engineering, which upgrades prompt engineering into building automated loop systems, providing multiple production-grade patterns and tools, and supporting Claude Code, Codex, etc.
The article introduces the Agent Plugins 1.0.0 specification, which aims to solve the binding and distribution problems between Agent Skills and MCP servers through a standardized portable directory format, and has been adopted by Google, Vercel, and several other companies.
Google Cloud's Agent Plugin lets developers package skills, MCP servers, and tools into a single portable folder, enabling build-once-use-everywhere workflows for AI agents.
OJCP (Open Job Context Protocol) is a new open standard built on MCP that lets AI agents discover, evaluate, and apply to jobs in a structured, consent-first way. It defines schemas and MCP tools for job search, application, and status tracking, aiming to standardize agent-job interactions.
Pillar Security researchers detail Deadbugz, an active supply-chain campaign that distributes a malicious MCP server through GitHub pull requests, gating credential-stealing instructions behind three benign tool calls.