Denoise · Twitter

The conversation is shifting from theoretical exploits to applied security engineering, with Anthropic and GoogleDeepMind leading the public discourse on agent safety.

Major labs are publicly releasing agent red-teaming frameworks and disclosing vulnerabilities, while indie researchers test them in the wild.

• 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 primary developer experience for AI coding is fragmenting from IDE-centric tools like Copilot towards terminal-native agents from actors like Anthropic.

Anthropic released Claude Code 1.5, a terminal-native agent, prompting benchmarks and discussion about a fundamental shift in coding workflows.

• 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 new stack for agent orchestration is converging, with OpenAI and Anthropic offering competing protocol primitives while platforms like Vercel provide the compute layer.

OpenAI released a new agent SDK, while Vercel and Replit shipped agent deployment runtimes, showing a focus on production infrastructure.

• 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 model capabilities get headlines, major players like MistralAI continue to commoditize progress by releasing foundational, high-quality datasets.

MistralAI released a large, open Optical Character Recognition (OCR) dataset intended for training multimodal models.

• Mistral AI@MistralAIrising

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

The simple RAG pattern is under pressure, with practitioners like Greg Kamradt and startups like mem0ai advocating for more complex, stateful memory systems.

Developers are stress-testing massive context windows and finding failure modes, leading to calls for more advanced 'context engineering' beyond RAG.

• 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 principles of agentic workflow automation, led by AI-native companies, are being adopted by established SaaS players like Notion and Linear.

Workspace automation tools like Notion and Linear are adding AI-driven features, mirroring patterns seen in agentic systems.

• 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 a craft of individual 'tricks' into a systematic, data-driven optimization problem, led by platforms like Weights & Biases.

The focus is on systematic, large-scale benchmarking of system prompts to find optimal performance patterns across different models.

• 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 scales, the bottleneck is shifting from raw data volume to sophisticated filtering and curation of synthetic datasets, as highlighted by @jerryjliu0.

A discussion highlights the importance of carefully curating synthetic data for training agents to avoid poisoning generalization.

• Jerry Liu@jerryjliu0repeated

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