Denoise · Twitter

The security focus is shifting from model-level prompt injection to system-level vulnerabilities in agent orchestration and tool interaction, as seen in real-world tests by MalwareTechBlog.

Major labs like Anthropic and Google DeepMind are publicly disclosing agent red-teaming frameworks and patched vulnerabilities.

• 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 developer workflow is being debated: the integrated IDE (Cursor + Copilot) versus the conversational terminal agent (Claude Code), with early adopters like levelsio claiming higher productivity on the latter.

Anthropic released Claude Code 1.5, a terminal-native agent, prompting immediate benchmarks and developer workflow re-evaluations.

• 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.

The stack is converging on managed agent workers and orchestration primitives. OpenAI's SDK and Vercel's edge runtime both provide solutions for deploying durable, stateful agents, abstracting away the underlying complexity.

OpenAI, Vercel, and Replit all launched new infrastructure for deploying and orchestrating 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.

While agents dominate discourse, foundational work on data continues, with major labs like MistralAI providing the cleaned, large-scale datasets necessary for future model improvements.

MistralAI released a large-scale, cleaned web OCR dataset to support development of multimodal models.

• Mistral AI@MistralAIrising

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

The consensus is that vector search is not a complete memory solution. Startups like mem0ai and frameworks like LlamaIndex are now focused on hybrid approaches, differentiating between working memory and long-term knowledge graphs.

Discussions center on the limitations of simple RAG and the move towards more sophisticated 'context engineering' frameworks.

• 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 pattern of durable, replayable workflows is converging. Temporal explicitly offers this for systems, while Notion and Linear are implicitly applying the same automation principles to human productivity tools.

Workspace automation is becoming a standard feature, with Notion and Linear launching tools for auto-filling and auto-triaging.

• 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 bifurcating between practitioner wisdom (dotey's thread) and industrial-scale analysis (Weights & Biases' 40k prompt benchmark), showing a formalization of the discipline.

Prompt engineering is maturing from anecdotal tricks to systematic, large-scale benchmarking and optimization.

• 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 model capabilities plateau, data quality becomes the key differentiator. Jerry Liu's work highlights the emerging challenge of identifying and removing synthetic data that harms generalization.

The focus for training effective agents is shifting to high-quality data curation and filtering of synthetic data.

• Jerry Liu@jerryjliu0repeated

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