Denoise · Twitter

@GoogleDeepMind and @AnthropicAI are converging on the idea that agent security vulnerabilities lie in orchestration and tool interaction, not just input sanitization.

Attention is focused on creating formal frameworks for red teaming autonomous agents, moving beyond simple prompt security.

• 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 is a direct implementation of the workflow shift that @karpathy described, challenging incumbent IDE-based tools.

Major releases and endorsements signal a fundamental shift in developer experience towards terminal-native coding agents.

• 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 clear layering is emerging: @OpenAI is defining the protocol, @LangChainAI provides integration glue, and @vercel and @replit are building the edge runtimes.

The infrastructure stack for deploying and orchestrating agents is rapidly being built out by major platform 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 dataset release indicates that foundational data curation for vision tasks remains a priority, even as agentic systems dominate the discourse.

The primary signal was a large-scale public dataset release for web-based optical character recognition (OCR).

• Mistral AI@MistralAIrising

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

Actors like @GregKamradt and @mem0ai are converging on the idea that vector search is an insufficient primitive for agent memory, pushing for more structured approaches.

The conversation is evolving from simple RAG to more sophisticated 'context engineering' and complex agent memory systems.

• 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.

The auto-triage from @linear and chained updates from @NotionHQ mirror the pattern of autonomous task completion seen in more explicit AI agents.

Workspace productivity tools are shipping agent-like automation features for tasks like issue triage and 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.

@weights_biases's large-scale benchmark refutes the value of small, isolated prompt experiments, pushing the community towards more rigorous, data-driven methods.

Efforts are shifting from anecdotal prompt 'tricks' to large-scale, systematic benchmarking to find optimal system prompts.

• 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 analysis reveals that naive use of synthetic data can poison generalization, making sophisticated filtering a critical step in the MLOps pipeline for agents.

The focus is on the nuances of dataset curation for training agents, specifically filtering harmful synthetic data.

• Jerry Liu@jerryjliu0repeated

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