Fetching from the wire…
01
02
03
04
05
06
07
08
Tools2026-09-04 · source-backed
Buried in the bug list: goose now fails closed on invalid default GCP credentials, rejects cmd.exe commands containing newlines, sanitizes hidden Unicode in Bedrock tools, bounds non-streaming JSON responses, and gives an honest compaction failure message instead of reporting success when no tool responses exist. Each one is a case where the agent previously kept going with a broken precondition, which is the failure mode that costs a whole unattended run. The release also adds an on_failure block for PreToolUse hooks and a PreToolUseResult event with a stable tool_call_id across the tool lifecycle, promotes web-search and browser-use to built-in skills, and ships Linux ARM64 desktop packages. GitHub
Each link below shares sources, entities, or timing with this story.
AWS made the managed AgentCore harness generally available on June 18. You define model, tools, skills, and memory with CreateHarness, then run it with InvokeHarness. It ships multi-model support (Bedrock, OpenAI, Gemini, LiteLLM), mid-session context preservation, built-in br...
Block released it August 27 with a security section dominated by defaults that previously failed open: fail closed on malformed tool visibility, permission denies now take precedence, fail closed on invalid default GCP credentials and invalid Codex ACP mode, honor plugin enabl...
1. Set package cooldown to 72 hours across all your package managers. pnpm: resolution-time=72h, uv: --exclude-newer, npm via .npmrc. This single config change would have protected you from the LiteLLM attack. Willison's survey covers all seven managers. 2. Install Lasso Secur...
Microsoft Threat Intelligence disclosed ChainDrop on August 4: a self-propagating npm worm that poisoned 444 packages across 2,212 versions in under four hours, starting from [redacted] at 150M weekly downloads, plus flat-cache and file-entry-cache. Corroborated by Unit 42, St...
1. Deploy Nemotron 3 Super for Agentic Reasoning (ml-ops, advanced) — 120B MoE activating only 12B params. vLLM with --reasoning-parser nemotron_v3. NVIDIA Blog 2. Build Multimodal RAG with Gemini Embedding 2 (ml-ops, intermediate) — Text, images, video, audio in one 3072-dim...
Buried at 4 points on Show HN, which is a shame, because it's the most consequential idea in today's Show HN cluster. Every MCP server you connect burns context on tool schemas before the agent does any work, and aggregation behind a two-tool facade is the obvious fix nobody h...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.