AI Agent News: Releases, Launches and Usage Jumps, Live · page 28
-
Strix passes 70K GitHub stars
-
Xerj passes 10K GitHub stars
-
Anatomy gained 10.3K GitHub stars in a week
Now at 17.1K stars.
-
AutoSwarm-Shadow-Adapter gained 1,510 GitHub stars in a week
Now at 1,783 stars, against about 47 a week over the month before.
-
PentAGI gained 4,041 GitHub stars in a week
Now at 31.3K stars, against about 771 a week over the month before.
-
Rea gained 166K GitHub stars in a week
Now at 339K stars, against about 43K a week over the month before.
-
Show HN: WhatsApp MCP running free on Cloudflare
WhatsApp MCP server implementation that runs on Cloudflare's infrastructure at no cost.
-
Hybrid Reasoning Agents: LLM agents that think in latent space, act through text
Research on LLM agents that perform reasoning in latent space while taking actions through text.
-
AI Gateway logs now show session affinity
Vercel's AI Gateway now includes a Sessions view that groups requests by session and displays session affinity routing, costs, models, providers, and cache metrics for coding agents and apps.
-
Graphify post checked: 128.3k GitHub stars, YC backing per its site, code maps without an LLM
A viral X post called Graphify a YC-backed tool with 127k stars. GitHub shows 128.3k, and Graphify's site says it is backed by Y Combinator (as reported). The package name graphifyy is right: PyPI lists it with a double y. Code is parsed locally without an LLM; docs, PDFs, images and video still go through your assistant's model.
-
Websites are closing their doors to AI agents
Websites are implementing measures to restrict access by AI agents.
-
Juicebox pricing changed: Free + $99/mo → Free + $119/mo
Free plan with limited searches; Starter at $119 and Growth at $199 per seat/month; Business custom plan; Juicebox Agents add-on at $199 per agent/month.
-
OpenCode has a security breach last that impacted 1.4% of console users
OpenCode experienced a security breach affecting 1.4% of console users.
-
Rea 6.7.0 released
Client configuration edits follow the duplicate-member semantics used by MCP clients.; Includes the 6.6.0 changes and migration notes. That version was tagged but was not published to npm.; Client configuration: Setup edits the last repeated JSON/JSONC…
-
LiteLLM 1.106.0 released
The cosign claims were validated; The signatures were verified against the specified public key; test(mcp): fire MCP client test deadlines on conditions instead of wall-clock time
-
Oh My Pi downloads up 34% in a week
176K npm and PyPI downloads in the 7 days to Oct 18, against 131K the week before.
-
PI Mcp Extension downloads up 46% in a week
105K npm and PyPI downloads in the 7 days to Oct 18, against 71.9K the week before.
-
AI Memory gained 1,575 GitHub stars in a week
Now at 10.7K stars, against about 616 a week over the month before.
-
Anatomy gained 12.2K GitHub stars in a week
Now at 16.3K stars.
-
Esp32 Iphone Agent gained 3,136 GitHub stars in a week
Now at 3,175 stars.
-
OpenCode gained 4,343 GitHub stars in a week
Now at 216K stars, against about 1,627 a week over the month before.
-
SPACE (Perplexity) gained 1,576 GitHub stars in a week
Now at 1,584 stars.
-
Rea gained 207K GitHub stars in a week
Now at 326K stars, against about 29.5K a week over the month before.
-
ComfyUI node cuts MiniMax H3 character video to about 72 seconds per clip, as reported
ComfyUI-Ref2VA-VSA uses sparse attention to render a 5.16-second, 1344×768 clip in about 72 seconds on an RTX 4090, 9x faster than native H3 (as reported by the repo). That is still about 14 seconds of compute per second of video by our calculation, not real time, and the README asks for a 24GB GPU.
-
Junie 3844.1 released
(+1 more release that day)
-
DeepSeek V4.1 Flash 1017 now available on AI Gateway
DeepSeek V4.1 Flash 1017 is now available on Vercel's AI Gateway with improvements to reasoning, coding, agentic tasks, and image understanding.
-
Context7 downloads down 32% in a week
498K npm and PyPI downloads in the 7 days to Oct 17, against 732K the week before.
-
Browser Harness downloads up 41% in a week
2.3M npm and PyPI downloads in the 7 days to Oct 17, against 1.6M the week before.
-
Hermes Paperclip Adapter downloads up 58% in a week
21.8K npm and PyPI downloads in the 7 days to Oct 17, against 13.8K the week before.
-
Kimi Code downloads up 27% in a week
40.5K npm and PyPI downloads in the 7 days to Oct 17, against 31.9K the week before.
-
OpenShell downloads up 36% in a week
129K npm and PyPI downloads in the 7 days to Oct 17, against 95K the week before.
-
Paperclip downloads up 26% in a week
39.3K npm and PyPI downloads in the 7 days to Oct 17, against 31.1K the week before.
-
PI Mcp Extension downloads up 44% in a week
99.6K npm and PyPI downloads in the 7 days to Oct 17, against 69.1K the week before.
-
TesterArmy e2e downloads up 48% in a week
236K npm and PyPI downloads in the 7 days to Oct 17, against 159K the week before.
-
AI Memory passes 10K GitHub stars
-
Corsair gained 1,616 GitHub stars in a week
Now at 15.1K stars, against about 642 a week over the month before.
-
Codex Security 0.4.0 released
Hosts can run Standard repository or path scans through the TypeScript runHostedScan(input, { executor, signal, onEvent }) API or the codex-security scan --host JSON-RPC adapter. Hosted execution…; Scan token totals include recorded usage from earlier Deep…
-
Building trust in AI agents: A production-ready evaluation framework with Amazon Bedrock AgentCore
AWS published an evaluation framework built on Amazon Bedrock AgentCore for detecting quality regressions in AI agents across development and production stages.
-
Warp: Scheduled cloud agents unavailable
Major impact; resolved after 2 h 34 min.
-
Introducing Reflection Beam: an open-weight model for agentic coding on AWS
Reflection AI's Beam, an open-weight model for agentic coding, can be deployed on AWS using SageMaker JumpStart or HyperPod Inference with various inference frameworks.
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.