Denoise · Twitter

The attack surface has moved from model-level prompt injection to vulnerabilities in the agent orchestration and tool-use layers, a focus for both Anthropic and Google DeepMind.

Security research is now centered on red-teaming autonomous agents, with new frameworks from Google DeepMind and real-world pentesting results being shared.

• Anthropic@AnthropicAIrising

  Responsible disclosure on a Claude jailbreak chain we patched last week. Full write-up including our red team timeline.

• Google DeepMind@GoogleDeepMindrising

  New red team framework for prompt injection in autonomous agents. Covers cross-tool leakage, scanner evasion, and sandbox escape patterns.

• MalwareTech@MalwareTechBlogrepeated

  Autonomous agent running pentest flows against a real SaaS. First real-world run: fewer false positives than I expected on the vulnerability surface.

A new developer experience philosophy is emerging, with players like Anthropic and voices like Karpathy betting on a terminal-centric future over GUI-based tools like Cursor.

Anthropic's Claude Code 1.5 launch marks a major push toward terminal-native coding agents, with early benchmarks and developer testimonials indicating strong uptake.

• Anthropic@AnthropicAIrising

  Claude Code 1.5 is live. Terminal-native coding agent with full Claude Opus reasoning, file-ops sandbox, and session replay.

• Andrej Karpathy@karpathyrising

  The developer-experience shift from IDE to terminal agent is underrated. Coding workflows are about to look nothing like 2024.

• swyx@swyxrising

  Codex vs Claude Code terminal agent benchmarks. Pass@1 diverges more than I expected on the long-context editor tasks.

• DSPy@dspy_airising

  DSPy 3.0: prompt optimization via compile-time search over system prompt variations. Benchmarks inside.

• @levelsio@levelsiorising

  Switched my whole editor setup to Claude Code this week. Shipping faster than when I used Cursor + Copilot.

The market is converging on a standard architecture for agents: an orchestration layer (OpenAI SDK), a durable runtime (Vercel, Replit, Temporal), and a common model interaction protocol.

A wave of new infrastructure for AI agents has been released, including an SDK from OpenAI and dedicated deployment platforms from Vercel and Replit.

• OpenAI@OpenAIrising

  New agent SDK: protocol-level tool calling, deployment harness, and multi-worker orchestration primitives. Docs live.

• LangChain@LangChainAIrising

  MCP protocol integration thread. How to wire existing LangGraph agents into the Anthropic Model Context Protocol server spec.

• Vercel@vercelrising

  Edge runtime for agent workers is live. Spawn durable background agents from any serverless deployment.

• Alex Albert@AlexAlbert__rising

  When your security scanner finds nothing scary on an agent deploy, check the orchestration layer again. That's usually where the jailbreak sneaks through.

• Replit@replitrising

  New agent deployment harness. One command to go from local orchestration to hosted agent worker.

While the agent stack dominates today's conversation, MistralAI's release shows that access to high-quality, large-scale training data remains a key competitive vector for foundational models.

MistralAI released a massive, cleaned 100M-row web OCR dataset, continuing its strategy of contributing foundational open data assets.

• Mistral AI@MistralAIrising

  Open dataset release: 100M-row web OCR dataset. Cleaned, licensed, ready to train.

The term "context engineering," popularized by Greg Kamradt, is gaining traction, framing a move towards hybrid systems that combine massive context, strategic caching, and structured memory from mem0ai.

The conversation is evolving from simple RAG to managing extremely large context windows and engineering sophisticated, multi-layered memory systems for agents.

• Vaibhav Srivastav@reach_vbrising

  Tested the new 10M context memory window end to end. Surprising failure modes around rag retrieval cache invalidation, thread below.

• Greg Kamradt@GregKamradtrising

  RAG is dead, long live context engineering. My framework for when to cache, when to retrieve, and when to just dump memory into the prompt.

• mem0@mem0airising

  Memory layer for agents: differentiating working memory from the subconscious store. Vector index isn't enough anymore.

• LlamaIndex@llamaindexrepeated

  Knowledge graph retrieval walkthrough: when semantic vector search misses, graph hop beats it every time.

Core principles of agentic systems—durable execution, state management, automated triage—are being integrated into mainstream productivity tools, signaling broader adoption beyond the core AI developer.

Workspace automation features are becoming standard in SaaS tools like Notion and Linear, while Temporal positions its workflow engine for agent orchestration.

• Notion@NotionHQrising

  Notion workspace automation is out of beta. Auto-fill tables, chained updates across databases, and a new audit log surface.

• Linear@linearrising

  Linear now auto-triages incoming issues. Quiet launch, but already our favorite workspace feature of the year.

• Temporal@temporaliorepeated

  Orchestrating agents with durable workflows: replayable, resumable, and multi-worker by default. Walkthrough from our infra team.

• James Clear@jamesclearrepeated

  The best habit tracker is the one you actually open. Three open-source alternatives worth trying.

Prompt engineering is being formalized through systematic benchmarking at scale (Weights & Biases) and automated optimization frameworks like DSPy 3.0.

Practitioners are sharing advanced prompting techniques and large-scale benchmark results, moving the practice from craft to a more systematic discipline.

• dotey@doteyrising

  Five prompt tricks learned this week from reviewing 200 production prompts. Short thread.

• Weights & Biases@weights_biasesrising

  System prompt benchmarking at scale: we ran 40k variants across 6 frontier models. The efficient frontier is not where you think.

As agent capabilities scale, the bottleneck returns to data quality. Jerry Liu's work highlights the critical problem of filtering synthetic data that looks plausible but poisons model generalization.

The focus in agent training infrastructure is shifting to the data pipeline, specifically the challenge of curating high-quality synthetic data.

• Jerry Liu@jerryjliu0repeated

  Dataset curation for agent training: how we filter synthetic data that looks good but poisons generalization.