Countinghouse, an Apache-2.0 multi-tenant Model Context Protocol (MCP) runtime on npm, lets MCP tools call each other in-process rather than through a model, reviving a decade-old design from the auth…
Snowflake announced the general availability of per-user AI credit quotas for CoCo and introduced three new governance controls—MDM, agent profiles, and restricted session scope—to manage access, beha…
A developer has created the Claude Task Completion Verifier, an MCP server designed to eliminate 'phantom completions' in AI agents by replacing self-reported success with deterministic verification. …
ChatGPT plugins and Claude connectors both now reliably handle multi-step calendar tasks, with ChatGPT's work tier able to chain actions like reading a week of events and inserting recovery blocks, ac…
Only 12% of companies are always confident in the factual accuracy of answers from their primary discovery tools, according to IDC research cited in an Elastic webinar, and over 50% of companies are s…
An engineer argues that agent permissions are becoming product design, not just security plumbing. The piece highlights the Hermes Agent Blank Slate setup, which starts with a minimal agent and requir…
Ayeixa has released MCP Guardian, an open-source middleware that adds capability fencing and parameter sanitization to Model Context Protocol servers, blocking directory traversal and dangerous shell …
Postman identifies four infrastructure layers—system of record, discovery, security, and observability—that production AI agents need to operate reliably, citing its Context Graph and MCP Server as so…
A developer details how to use Cline, an autonomous AI code agent for VS Code, safely in production by designing deliberate constraints. The developer argues that while Cline can write files, execute …
AgenticSchema, a new open-source library, converts existing Schema.org structured data (JSON-LD, microdata, RDFa) on web pages into Model Context Protocol (MCP) tools that AI agents can call, eliminat…
B2B SaaS companies in 2026 are moving away from per-task pricing models like Zapier's to avoid escalating costs, according to a blog post from Marketing So High. The post recommends evaluating alterna…
A developer building AgentShare, a platform for agent-ready API discovery and x402 payment rails, is seeking community input on which API niches AI agents will pay for with USDC first. The developer h…
MakePlans launched an MCP server for its appointment booking platform, enabling AI assistants like Claude and ChatGPT to manage bookings via natural language. The server uses the open Model Context Pr…
The Model Context Protocol (MCP) is moving toward a stateless protocol model, allowing requests to reach any server instance without sticky routing or shared session state. A developer recently migrat…
The Model Context Protocol (MCP) 2026-07-28 specification removes the stateful session model, eliminating the initialize handshake and Mcp-Session-Id header, requiring all requests to be self-containe…
A developer argues that small businesses should start with a draft queue workflow for MCP projects rather than autonomous agents, emphasizing human approval to reduce risk. The approach includes shado…
Rogue agent behaviour in AI deployments typically appears as legitimate-looking tool use that slips past permission checks, according to a practical guide covering assessment and defense strategies. T…
The Federal Reserve's SR 11-7 model risk management framework, built on deterministic model assumptions, breaks down for agentic AI systems that plan, call tools, and act through multi-step reasoning,…
A new arXiv study, 'Exposed by Design,' found more than 21,000 Model Context Protocol (MCP) server instances reachable from the public internet, with 91.8% of audited servers lacking OAuth and 687 too…
An engineer at MCP explains the key differences between the stdio and Streamable HTTP transports for the Model Context Protocol, noting that the HTTP+SSE transport from the 2024-11-05 revision has bee…