Denoise · Twitter

The discourse is moving from theoretical exploits to structured disclosure and testing, with @AnthropicAI and @GoogleDeepMind leading formalization while practitioners like @MalwareTechBlog test viability.

Major AI labs are establishing public norms for agent security, with Anthropic disclosing a patched jailbreak and DeepMind releasing a red team framework.

• 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.

The focus has shifted from IDE plugins (Copilot) to standalone agents (Claude Code), with @karpathy framing it as a major platform transition and @swyx providing early comparative benchmarks.

Anthropic's Claude Code 1.5 release of a terminal-native agent is driving conversations 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 race is on to provide the defining protocol and runtime for agents, with OpenAI's SDK, Vercel's edge workers, and LangChain's protocol integrations all converging on the orchestration problem.

Major infrastructure providers including OpenAI, Vercel, and Replit released new primitives for deploying and orchestrating autonomous agents.

• 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.

The primary signal is a large-scale data release from MistralAI, which serves as a foundational building block for future model development rather than a direct product release.

MistralAI released a 100M-row web OCR dataset, providing a significant new resource for training multimodal models.

• Mistral AI@MistralAIrising

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

The community, led by voices like @GregKamradt, is moving past the term RAG to describe a new set of problems involving caching, memory hierarchies (@mem0ai), and graph retrieval (@llamaindex) in massive contexts.

With 10M+ token context windows being tested, the conversation is shifting from simple RAG to more complex "context engineering" and memory architectures.

• 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.

SaaS tools like Notion and Linear are building agent-like automation features, while platforms like Temporal offer the robust, stateful orchestration primitives required to build them from scratch.

Workspace automation sees new releases from Notion and Linear, with underlying durable execution patterns being discussed by infrastructure providers like Temporal.

• 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 maturing from an art to a data-driven science, with platforms like @weights_biases enabling quantitative analysis, complementing qualitative practitioner insights from people like @dotey.

The focus is on systematizing prompt engineering, with large-scale benchmark results from Weights & Biases and practical tips from production reviews.

• 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 becomes more accessible, the bottleneck is shifting from compute to high-quality data, with a focus on sophisticated filtering for synthetic data as highlighted by @jerryjliu0.

The discussion centered on the nuances of dataset curation for training agents, specifically the challenge of filtering out synthetic data that harms generalization.

• Jerry Liu@jerryjliu0repeated

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