Denoise · Twitter

The security focus is expanding from model-level jailbreaks (Anthropic) to vulnerabilities in the agent's orchestration and tool-use layer (Google DeepMind, @MalwareTechBlog).

Major labs are releasing frameworks and disclosures for red-teaming autonomous agents, focusing on prompt injection and tool interaction 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.

The primary developer interface may be shifting from IDE plugins like Copilot to standalone terminal agents like Claude Code, a trend observed by @karpathy and early adopters.

Anthropic's release of Claude Code 1.5, a terminal-native agent, has triggered a wave of adoption and discussion about the decline of traditional IDE 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 convergence is happening around agent orchestration, with OpenAI's SDK and Vercel's edge runtime providing competing but similar solutions for deploying durable agent workers.

OpenAI, Vercel, and Replit all released primitives for agent deployment, signaling a race to build the standard infrastructure for hosting and managing 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 the conversation, foundational model providers like Mistral AI continue to invest in building core data assets for multimodal capabilities.

Mistral AI released a large, cleaned web OCR dataset to facilitate the training of new multimodal models.

• Mistral AI@MistralAIrising

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

A consensus is emerging that vector search alone is insufficient for agents, leading to new architectures from @mem0ai and frameworks from @GregKamradt.

The conversation is shifting from simple RAG to more sophisticated 'context engineering' with complex memory stores and graph-based retrieval.

• 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 architectural patterns for agents, such as durable orchestration from Temporal, are being productized as features in mainstream productivity software.

SaaS tools like Notion and Linear are shipping agent-like automation features directly into their products, such as auto-filling tables and triaging issues.

• 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 practice is maturing from anecdotal tricks to data-driven optimization, with platforms like Weights & Biases running massive experiments to find the efficient frontier for system prompts.

Efforts are underway to make prompt engineering more systematic through large-scale benchmarking and sharing of production-tested techniques.

• 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 complexity increases, the bottleneck returns to data quality. @jerryjliu0 highlights the critical need for better filters for synthetic training data to avoid poor generalization.

The focus in agent training infrastructure is on the difficult problem of curating high-quality datasets and filtering out harmful synthetic data.

• Jerry Liu@jerryjliu0repeated

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