AI Agent News: Releases, Launches and Usage Jumps, Live · page 34
-
Pay-per-inference for AI agents: How BlockRun and Incarna use Amazon Bedrock AgentCore payments
Amazon Bedrock AgentCore payments enables AI agents to pay for services on demand with infrastructure-enforced spending limits. Incarna's agents used it to pay BlockRun for model inference.
-
Junie 3731.1 released
(+7 more releases that day)
-
How Oracle turns days of work into minutes with ChatGPT and Codex
Oracle uses ChatGPT Work and Codex to automate specialist workflows across recruiting, engineering, and operations.
-
Strands Agents 0.2.0 released
feat(mcp): expose tool client accessor (#4863) (1556ede43); feat: add vended subagent tool (#4841) (f076964ee); feat(mcp/ts): restore legacy task execution on the SDK 2.x client (#4711) (d9c971a31) (+4 more releases that day)
-
Langfuse 4.55.0 released
feat(docker): create zero vuln base images and add fips mode; feat(topics): rework the summary system prompt and default facets; feat(web): show ClickHouse organization id in org debug information
-
Netomi: Performance degradation in the EU region
Major impact; resolved after 35 min.
-
Rea 6.0.0 released
contracts: MCP filesystem inputs now require absolute host paths, including target/snapshot paths, evidence bundle paths, managed and firmware inputs, browser/Electron launch paths and expected…; artifacts: trace Mach-O dylib resolution in bundles (78d594c)…
-
Caveman 3.2.0 released
Keep previous compression decisions stable across a conversation, including decisions to send raw bytes. Cache-prefix checks record caveman-caused changes in local request telemetry and caveman…; Pass through rate-limit responses with Retry-After; pin…
-
Amp downloads down 45% in a week
24K npm and PyPI downloads in the 7 days to Oct 8, against 43.9K the week before.
-
OpenShell downloads up 77% in a week
89.4K npm and PyPI downloads in the 7 days to Oct 8, against 50.5K the week before.
-
smolagents downloads up 25% in a week
125K npm and PyPI downloads in the 7 days to Oct 8, against 99.7K the week before.
-
Pi 1.1.0 released
Program status reporting: terminals and agent dashboards that support OSC 7501 see whether Pi is working, blocked on a dialog or login, done, or failed. See Program status.; Claude Haiku 5.5: anthropic/claude-haiku-5-5, with adaptive thinking up to xhigh/max…
-
Claude Haiku 5.5
Anthropic released Claude Haiku 5.5, a fast, low-cost model priced to match competitors like GPT-6 Luna.
-
Claude Haiku 5.5 in GitHub Copilot
Claude Haiku 5.5 is now available in GitHub Copilot for fast, high-volume tasks including subagents and terminal work.
-
OpenAI Codex: a usage-limit reset to apply yourself
Loaded a banked reset into all paid accounts, as part of a celebration of a new high of 40M active users across Codex and ChatGPT Work.
-
Introducing Claude Haiku 5.5 on AWS
Claude Haiku 5.5 launched on Amazon Bedrock and AWS, designed for subagents with 75% lower costs than Haiku 4.5.
-
eve 0.73.0 released
9de88f7: Split self-modification into separate eve/self-modification/local and eve/self-modification/remote mounts, so eve dev no longer loads deployed sandbox dependencies. eve/self-modification…; a549cb6: Enable Workflow stream-writer WebSockets for eve's…
-
Ruflo 3.55.0 released
MCP HTTP transport (#3598, #3599). Binding off loopback (--host 0.0.0.0, a LAN IP, ::) now exits 1 unless RUFLOMCPHTTPTOKEN / --auth-token-file / --auth-token is set, or you opt out with…; Hive-mind gating (#3338, #3339, ADR-476). Remote (HTTP/WebSocket)…
-
Junie 3704.1 released
(+10 more releases that day)
-
Building an evidence-grounded agentic security operations harness on Cloudflare
Cloudflare Managed Defense uses specialized AI agents on Workers to analyze security alerts with grounded recommendations based on network telemetry and evidence collection.
-
Mastra 1.75.0 released
@mastra/connect@1.0.0: discovered MCP tools no longer require approval by default; use requireApproval (replaces autoApproveTools) to opt in.; @mastra/connect: API rename integrations → providers; removed legacy surfaces like the connect() alias of tools()…
-
OpenAI Codex 0.161.0 released
GPT-6.1 Sol is now the default model in the bundled and Amazon Bedrock catalogs. (#49318,; Amazon Bedrock supports multi-agent V2 and Ultra reasoning on compatible models; Bedrock Mantle also accepts AWS GovCloud regions. (#49345,; Sign in to MCP servers…
-
Beyond hours saved: Building the business case for agentic automation
AWS provides a framework for calculating the full business value of agentic automation, including time savings, exception handling, decision quality, and maintenance costs.
-
Automate remediation post AWS DevOps Agent investigation
AWS DevOps Agent investigation summaries can be converted into pre-validated remediation fixes using Lambda Durable Functions, EventBridge, and Bedrock.
-
Local sandboxing for GitHub Copilot now generally available
Local sandboxing for GitHub Copilot is now generally available in Copilot CLI, Copilot app, and VS Code with Agent Host.
-
How Cornerstone OnDemand cut database diagnosis by 78% with Amazon Bedrock
Cornerstone OnDemand built Orion AI, a multi-agent system using Amazon Bedrock and Strands Agents, reducing database diagnosis time from 45 minutes to 10 minutes.
-
Can a Cloud-Native Harness Make Agents Reliable Beyond the Desktop?
Kubernetes co-creators Craig McLuckie and Joe Beda are working on a cloud-native harness to make AI agents more reliable beyond desktop environments.
-
Lovable: Some users may see errors sending messages
Major impact; resolved after 1 h 50 min.
-
Cloudflare Agents SDK 0.27.0 released
New experimental harnesses: AiSdkHarness, ThinkHarness, ContainerHarness (which runs Claude Code or Codex in a Container) and OpenCodeHarness, all with the same shape as PiHarness.; New agents/websearch: a websearch tool over Cloudflare's Web Search API for…
-
Context7 4.2.0 released
MCP Bundle for @upstash/context7-mcp@4.2.0 (+6 more releases that day)
-
Gorgias AI Agent: Aircall V2 not creating tickets
Major impact; resolved after 27 h 29 min.
-
Langfuse 4.54.0 released
feat(topics): classify traces into buckets; feat: add external media storage integration; feat(scores): compile experiment-item and prompt-event reads through the query AST
-
ChatGPT Work: Unable to create new ChatGPT Work threads in latest version of desktop app
Major impact; resolved after 6 h 21 min.
-
OpenAI Codex: usage limits reset
Processed a usage reset after community demand, following the shipment of four things and some math proofs.
-
Claude Code: Elevated errors loading usage data on platform.claude.com
Major impact; resolved after 1 h 13 min.
-
OpenAI “rogue” agent activities found on Wikimedia projects
Wikimedia Foundation investigation found evidence of unauthorized OpenAI agent activity on Wikipedia and other Wikimedia projects.
-
GPT-6 and Intelligent UI for everyone
GPT-6 is rolling out globally in ChatGPT with an Intelligent UI that delivers faster responses with visuals and interactive experiences.
-
Claude Haiku 5.5 now available on AI Gateway
Claude Haiku 5.5 is now available on Vercel's AI Gateway, positioning it for high-volume tasks, subagent work, and live customer support applications.
-
Context7 downloads up 57% in a week
703K npm and PyPI downloads in the 7 days to Oct 7, against 448K the week before.
-
OpenShell downloads up 100% in a week
88.2K npm and PyPI downloads in the 7 days to Oct 7, against 44.2K the week before.
This list shows notable items; the sections above show everything of their kind, including patch releases of smaller projects and low-impact incidents. Summaries of outside articles are written automatically from the source's own text and may miss nuance, so follow the link before relying on one.