Axios reports Morgan Stanley has begun blocking third-party agents, and JPMorgan, Bank of America and Citi are restricting them, after banks found agents using customers' own credentials and sometimes acting beyond user instructions. Most banks are moving toward read-only access, so if your agent touches finances, plan on open-banking APIs rather than a user's login.
Out of beta since Oct 22, the gpt 6 luna endpoint for yes/no, pick one and score questions adds prompt caching ($0.01 per 1M cached input tokens, $0.10 uncached, still no output charge), native JSON inputs, images in question instructions, US and EU residency, GPT Live integration, and up to 400 questions per request. Datadog says it scores the risk of agent generated shell commands before running them at 40x faster p95 than frontier models, which makes it a cheap gate for an agent's next step…
Released Oct 25, the fixes stop symlink and path tricks that let the filesystem server write or peek outside allowed directories (Windows only partly fixed), block other web origins from calling the everything server over HTTP, and close SSRF gaps in fetch via IDNA host names and NAT64 addresses. Package managers rank 2.0.1 below old date versions, so install it explicitly, e.g. npm install @modelcontextprotocol/server [email protected] , and update mcp server fetch to 2026.10.24…
The MIT licensed MCP server takes your agent's SQL and a short layout, runs each query read only against Postgres, SQLite, Excel or CSV, and draws charts, tables and headline numbers in a browser or a pane beside your terminal, with every number showing its query. Version 0.4.1 (Oct 22) adds a Claude desktop extension; install in Claude Code with /plugin install showboard marketplace rohanbeingsocial/showboard , and treat the author's 2.7x cost saving as a single dataset test (benchmark breakdown)…
The MIT licensed tool diffs .d.ts files and changelogs for JS/TS deps newer than your model's cutoff, writes ranked notes into AGENTS.md or CLAUDE.md , type checks after each edit, and exposes a search changes MCP tool for Claude Code, Cursor, Codex or OpenCode. It's early and maintained by one person; try it with npx stalemap (benchmark)…
The PUBG maker's AI team trained the 30B total, 4B active mixture of experts model from scratch for Korean and English chat and tool calling agents, and ships NVFP4 (vLLM, single GPU), MLX and GGUF builds alongside the main weights. It's a gated preview with automatic approval, so specs and results may change before the final release, and the hosted API and playground fall under separate terms…
Spotted something we missed? Start a thread.