Anthropic’s claude-api skill adds /claude-api build-eval (guided eval design in your repo) and /claude-api hillclimb (one-change-per-round tuning with a held-out set to catch overfitting). On an internal support bench, hillclimb lifted search accuracy from 74.4% to 98.9% while cutting cost to about one-fifth.
Announced Nov 10, the MCP lets assistants send SMS and RCS messages and manage contacts in a customer's TrueDialog account. It supports test sends and honors opt outs automatically…
Announced Nov 12, the server exposes 35 tools so Claude Code, Codex and Copilot CLI can create, validate, run and fix Lemma workflows. Agents act with the user's own Lemma permissions…
Wise's official MCP server lets assistants read balances, transfers, recipients and exchange rates through 15 read only tools. It can't move money; drafting transfers is listed as coming soon…
InterWorks' Curator 2026.11 02 ships an MCP server that lets agents build pages, upload dashboards and edit themes with token scoped permissions. Edits snapshot first so they can be rolled back…
Announced Nov 10, Userlytics MCP lets assistants query study data, AI Insights, transcripts and participant quotes with session recording links. Access is limited to the connected account's studies…
Spotted something we missed? Start a thread.
