Denoise · Twitter

The security focus is shifting from model-level safety to vulnerabilities in the agent orchestration and tool-use layers, a concern echoed by Anthropic, Google DeepMind, and practitioners.

Major labs are publicly disclosing agent jailbreaks and releasing red-teaming frameworks as agents begin to see real-world pentesting and deployment.

• 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 product category is maturing as Claude Code is directly benchmarked against OpenAI's Codex and GitHub's Copilot, signaling intense competition in agentic coding assistants.

Anthropic's release of Claude Code 1.5, a terminal-native agent, is driving a conversation about a fundamental shift in developer workflows away from traditional IDEs.

• 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.

A convergence on agent protocols is underway, with OpenAI and Anthropic releasing specs while frameworks like LangChain adapt to them and platforms like Vercel provide the compute fabric.

Major platforms including OpenAI, Vercel, and Replit are shipping new SDKs and runtimes for deploying and orchestrating agents, solidifying a new infrastructure layer.

• 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 conversation dominates, foundational dataset releases like Mistral's remain critical for enabling new model capabilities, particularly in document and image understanding.

Mistral AI released a large-scale, cleaned 100M-row web OCR dataset for public use in model training.

• Mistral AI@MistralAIrising

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

The dialogue is shifting from 'RAG vs. long context' to 'context engineering', a more nuanced approach combining caching, retrieval strategies, and memory architectures from actors like @GregKamradt and @mem0ai.

Developers are exploring new frameworks beyond simple RAG to manage state and memory in agents with massive context windows.

• 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.

Agentic automation principles are being integrated directly into SaaS products, as seen with Notion and Linear, to automate user workflows within their existing platforms.

Workspace automation features are becoming standard, with Notion and Linear releasing tools for auto-triaging and chained database updates.

• 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.

The art of prompt engineering is becoming a data-driven science, with platforms like Weights & Biases enabling optimization across thousands of variants, replacing anecdotal tricks with empirical evidence.

Practitioners and labs are moving towards systematic, large-scale benchmarking of system prompts to identify what works reliably across different models.

• 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 training scales, the key infrastructure challenge becomes less about raw data volume and more about building sophisticated pipelines for data curation, a point raised by @jerryjliu0.

The focus in agent training is shifting toward data quality, particularly techniques for curating and filtering synthetic data to avoid harming model generalization.

• Jerry Liu@jerryjliu0repeated

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