Denoise · Twitter

Agent security is rapidly maturing from ad-hoc vulnerability hunting into a formalized discipline, with @AnthropicAI and @GoogleDeepMind leading the conversation on responsible disclosure and systematic testing.

Major AI labs are now publicly disclosing agent jailbreaks and publishing formal red-teaming frameworks for autonomous systems.

• 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 consolidating around terminal-native agents, with @AnthropicAI's release validating the developer experience shift articulated by @karpathy.

Anthropic's release of a terminal-native agent, Claude Code 1.5, dominates the conversation, signaling a potential workflow shift 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 clear convergence on agent orchestration is visible, as @OpenAI, @LangChainAI, @vercel, and @replit are all shipping components for a common agent stack.

The ecosystem is rapidly building the infrastructure for agent deployment, with new SDKs, runtimes, and protocol integrations 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.

While agent infrastructure is the loudest signal today, @MistralAI's dataset release shows that foundational data curation remains a key, albeit quieter, driver of model progress.

MistralAI contributed a large-scale, open dataset for web OCR, providing a foundational resource for training multimodal models.

• Mistral AI@MistralAIrising

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

The simple retrieve-then-generate RAG pattern is being actively challenged by voices like @GregKamradt and @mem0ai, suggesting its limitations for advanced, stateful agents.

The discourse is evolving from simple RAG to 'context engineering,' focusing on cache invalidation in large contexts and sophisticated, multi-layered 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.

Agent-like capabilities are becoming table stakes in SaaS, with @NotionHQ and @linear shipping them, while infra players like @temporalio reframe their existing tech for the emerging agent market.

Workspace automation is becoming more sophisticated with agent-like features, while durable execution frameworks are being positioned 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 maturing into a data-driven discipline, with efforts like @weights_biases's massive benchmark run moving the field beyond anecdotal tricks.

The focus in prompt engineering is shifting towards systematic, large-scale benchmarking to discover optimal system prompt configurations.

• 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.

The single tweet from @jerryjliu0 on dataset curation points to a deeper, less visible challenge: ensuring the quality of synthetic data as agent training scales.

A specific technique for filtering synthetic data was shared, highlighting the importance of data curation for training effective agents.

• Jerry Liu@jerryjliu0repeated

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