Denoise · Twitter

The security conversation has matured from single prompt injections to complex, multi-step exploits in agentic systems, with Anthropic and Google DeepMind now formalizing these attack patterns.

Model providers and security researchers are publishing frameworks and disclosures on red-teaming autonomous agents, focusing on the orchestration layer.

• 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 ecosystem is bifurcating between end-user products like Claude Code and underlying frameworks like DSPy, which focuses on optimizing the prompts that power these agents.

Anthropic's Claude Code 1.5 release and commentary from developers like karpathy and swyx mark a clear move toward 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 stack is forming: OpenAI and LangChain are defining orchestration protocols, while Vercel and Replit are competing to provide the serverless runtime for agent workers.

Major infrastructure providers including OpenAI, Vercel, and Replit released new SDKs, runtimes, and deployment tools for orchestrating and hosting 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.

Instead of a new model, MistralAI's release focuses on the data layer, indicating that high-quality, specialized datasets are becoming a key competitive vector.

MistralAI released a 100M-row web OCR dataset, providing a foundational artifact for training future multimodal models.

• Mistral AI@MistralAIrising

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

A consensus is forming that a single vector index is insufficient for agent memory; mem0ai and GregKamradt are both advocating for more structured, hierarchical memory architectures.

Discussion shifts from simple RAG to 'context engineering,' exploring failure modes of large context windows and proposing multi-layered 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.

The patterns of agent orchestration, seen in devtools from OpenAI and Temporal, are now appearing as product features in general-purpose workspace tools like Notion and Linear.

SaaS tools like Notion and Linear are shipping autonomous features, such as auto-filling tables and triaging issues, reflecting broader agentic patterns.

• 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 field is moving beyond anecdotal prompt 'tricks' (@dotey) to a data-driven approach, where firms like Weights & Biases map the efficient frontier of prompt performance.

Prompt engineering is professionalizing through systematic, large-scale benchmarking of system prompts, as shown by work from Weights & Biases.

• 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.

Jerry Liu's commentary highlights a critical challenge: as agent training relies more on synthetic data, the core problem shifts from compute to building robust data filtering pipelines.

The focus in agent training infrastructure is on sophisticated data curation, specifically filtering synthetic data to avoid model degradation.

• Jerry Liu@jerryjliu0repeated

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