Topic · 115 stories · page 1 of 5
115 stories tagged agents on AI Tamer, newest first.
YC startup Pomerado mints deterministic MCP tools for sites with no API, including ones behind a login. MIT core runs locally with an OpenAI key.
Redwood's simulated control evals have an LLM fake tool outputs while an attacker works in real code. Safety error: 3.6 points, against 10.1 for BashArena.
Infinity saves an agent's conversation and stops it while a tool runs, then resumes it when the result arrives. AWS fits about 75,000 idle agents in 8 GB.
The open-source Rust CAD app now writes an MCP entry into Claude, Cursor, Codex, VS Code and six more AI clients when it starts. One setting turns it off.
The Rust GIS app, which its README says is mostly built with Claude, adds ACP agents, MCP over HTTP, run-scoped tokens, an audit log and per-agent undo.
The rollout on iOS and Android asks for six permissions, from Calendar to Home and Messages. The docs do not describe the feature yet, and early reports vary.
An agent asks, you approve in 1Password, and 1Password fills the login itself. Free on Individual and Family plans; most agent integrations are still coming.
Eight changes from our past week for people building with AI: model shutdowns, new defaults, two security fixes and new tools, each with a next step.
Walmart's Agent Gateway lets approved AI agents shop over MCP in a beta. Per-agent spending caps follow next month, and rule-breaking agents get blocked.
DuckDB 2.0.0 is out with Quack client-server, async I/O for S3, VARIANT, agent mode and triggers. Support runs to October 2027. Test before you upgrade.
Commerce.com stores now expose search and cart tools to in-browser agents via Cloudflare's edge WebMCP bridge, with a per-merchant opt-out rule.
OpenAI has acquired Orchestra Bio, a platform where AI agents track drug-development studies, and is folding its team into Rosalind. No terms were given.
Since 20 October, GPT-5.6 and later can hold a cached prefix for at least 12 hours in the Responses API. Writes cost 2.5x input; read prices are unchanged.
Exa 3.0 adds a larger embedding model, GA highlights that use 95% fewer tokens, billions of new documents and a $4 Lite search tier. Benchmarks self-run.
Modal's first Vera CPU server matched x86 on quiet single tasks but kept its pace as sandboxes piled up, doubling throughput in Modal's own early benchmark.
FirstMark Growth led a $60M Series B at a $1B valuation for Daytona, which rents sandboxes to AI agents. It says revenue grew about 25-fold this year.
rmcp 3.6.0 enforces the 2026-07-28 rule that servers must not send requests to clients, cancels abandoned calls, and bounds the stdio line buffer.
GitHub may require sign-in for some anonymous HTTPS clones and fetches of public repos during unusual traffic. SSH, pushes and authenticated Git still work.
Editing Copilot Studio agents inside Teams ends in January 2027 and the app leaves Teams in February. Agent management moves to the Copilot Studio web app.
MiniMax launched M3.1 Flash at $0.15/$0.60 per million tokens with a 1M context. M Plan requests to M2.5 and M2.7 now route to it; both retire 18 November.
Harness², a Google Research paper and Apache-2.0 repo from 9 October, refines an agent's harness per task with model weights fixed. Gains are the authors'.
In a 15 October Vals study, six assistant agents chased a refund past its deadline. One emailed an MP's office 50 times. Another used a private illness.
NVIDIA App 11.0.9.565 runs a local MCP server so AI agents can launch games and change settings. VideoCardz found it on by default. One toggle disables it.
Monty 1.2.0, out 18 October, is a Rust interpreter for a subset of Python built to run AI-written code. It adds match, csv, tomllib and hashlib.
