Hermes Agent Ships Tool Search for MCP: Anthropic Evals Show 49% to 74% Accuracy Gain on Opus 4

The open-source project adds local persistent memory to Hermes Agent through six layers, gated retrieval, and a wiki. The post Meet Memory OS: A 6-Layer Open-Source Memory Stack Built on Top of Hermes Agent appeared first on MarkTechPost.

Enterprises using the lightweight, open-source Flowise platform to power self-hosted AI workloads now have a new near-max-severity issue to worry about. Researchers at Obsidian Security have detailed a one-click remote code execution (RCE) vulnerability affecting self-hosted Flowise deployments through its implementation of Model Context Protocol (MCP) stdio servers. The problem is essentially a sandboxing failure of attacker-controlled MCP configurations, leading to server-side code execution. “Post-auth RCE in Flowise can be triggered with a single click via a malicious chatflow import before any save or run,” the researchers said in a blog post. “The official patch relies on input validation that is trivially bypassed and fails to address the root cause.” Flowise is commonly used to develop internal AI assistants, retrieval-augmented generation (RAG) applications, customer-facing chatbots, and autonomous agents connected to business systems. The flaw does not affect Flowise Cloud, a

Base launches Base MCP, letting ChatGPT and Claude agents connect to Base Accounts for swaps, transfers, portfolios, and app access. The post Base launches MCP to connect ChatGPT and Claude agents to onchain wallet actions appeared first on Crypto Briefing.

OmniVoice Studio runs voice cloning, video dubbing, real-time dictation, and speaker diarization entirely on your own hardware. No API keys, no cloud account, and no subscription required. The project supports 646 languages for TTS and exposes an MCP server for integration with Claude, Cursor, or any MCP client. The post Meet OmniVoice Studio: A Local, Open-Source Alternative to ElevenLabs appeared first on MarkTechPost.

There’s no denying the excitement around Model Context Protocol (MCP), an open protocol for connecting AI assistants with external data, tools, and APIs. Since its debut by Anthropic in late 2024, thousands of MCP servers have emerged for devops, cloud, and beyond. Now that developers have integrated MCP servers into applications, and they have been battle-tested, usage patterns are emerging. For instance, supplying better context for AI is the most commonly cited primary value of using MCP, according to Zuplo’s State of MCP report released in early 2026. The Zuplo report also found that 63% of MCP users adopt MCP servers for accessing data sources such as documentation or knowledge bases. In software development, context engineering is the act of supplying AI coding agents with relevant data and capabilities to improve the accuracy and relevance of their outputs. It also involves optimizing the breadth of information to guide efficient processing. Such context can include coding style

Tencent has open-sourced TencentDB Agent Memory, a fully local memory system for AI agents released under the MIT license. The project pairs symbolic short-term memory, which offloads verbose tool logs into a compact Mermaid task canvas, with a 4-tier long-term memory pyramid (L0 Conversation → L1 Atom → L2 Scenario → L3 Persona). It ships as an OpenClaw plugin and a Hermes Docker image, runs on local SQLite + sqlite-vec by default, and uses hybrid BM25 + vector retrieval with RRF fusion. Tencent's own benchmarks report a 61.38% token reduction and 51.52% relative pass-rate gain on WideSearch with OpenClaw, alongside PersonaMem accuracy moving from 48% to 76%. The post Tencent Open-Sources TencentDB Agent Memory: A 4-Tier Local Memory Pipeline for AI Agents appeared first on MarkTechPost.

Nous Research releases Contrastive Neuron Attribution (CNA), a method that identifies and ablates sparse MLP neuron circuits to steer LLM behavior — no sparse autoencoder training, no weight modification, and no degradation of general capability benchmarks. The post Nous Research Releases Contrastive Neuron Attribution (CNA): Sparse MLP Circuit Steering Without SAE Training or Weight Modification appeared first on MarkTechPost.

Perplexity has open-sourced Bumblebee, an internal security tool it uses to protect the developer systems behind its search product, Comet, and Computer. Bumblebee is a read-only inventory collector for macOS and Linux developer endpoints. It scans npm, PyPI, Go modules, MCP configs, editor extensions, and browser extensions — without invoking any package manager or running any code. The post Perplexity Open-Sources Bumblebee: A Read-Only Supply-Chain Scanner for Developer Endpoints appeared first on MarkTechPost.