Denoise · Twitter

The primary attack surface is moving from prompt injection to the orchestration and tool-calling layers, a vulnerability space defined by @GoogleDeepMind and @AnthropicAI.

Security research is shifting to autonomous agents, with major labs releasing red-teaming frameworks and disclosures on patched jailbreaks.

• 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.

@AnthropicAI's Claude Code release directly validates @karpathy's thesis of an IDE-to-terminal shift, fragmenting the previous dominance of tools like Copilot.

Major releases this week focus on terminal-native coding agents, representing a significant workflow shift away from traditional IDE integrations.

• 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 between @OpenAI, @AnthropicAI, and infra providers like @vercel and @replit to define the standard protocol and deployment harness for agents.

The infrastructure for deploying and managing agents is rapidly maturing, with new SDKs, runtimes, and orchestration primitives from major players.

• 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.

@MistralAI's open dataset release continues their strategy of commoditizing foundational resources to compete with closed-source players.

A large-scale, clean OCR dataset has been released, aimed at improving the training data available for multimodal models.

• Mistral AI@MistralAIrising

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

The consensus view, articulated by @GregKamradt and @mem0ai, is that naive vector retrieval is insufficient for agentic workflows, requiring new memory architectures.

The discourse is moving beyond simple RAG towards more sophisticated 'context engineering' and structured 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.

A convergence pattern is visible as @NotionHQ and @linear both ship features for AI-powered, structured workflow automation like auto-triage.

Workspace automation tools are adding AI-driven features for task management and issue triage, moving beyond simple content generation.

• 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 approach to prompting is industrializing, moving from individual advice (@dotey) to large-scale empirical analysis platforms like @weights_biases.

The focus in prompting is shifting from artisanal tricks to systematic, large-scale benchmarking of system prompts to find empirical performance gains.

• 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.

@jerryjliu0's post highlights that as synthetic data becomes common for agent training, the critical skill is identifying data that poisons generalization.

The key challenge in training data for agents is now filtering and curating high-quality synthetic data, not just generating it.

• Jerry Liu@jerryjliu0repeated

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