Denoise · Twitter

The security conversation is escalating from model-level exploits to system-level vulnerabilities in agent orchestration, a risk highlighted by Anthropic, DeepMind, and @AlexAlbert__.

Major labs like Anthropic and DeepMind are publicly releasing red team frameworks and vulnerability disclosures specifically for autonomous agents.

• 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 competitive front for AI coding assistants is moving from IDE plugins like Copilot to standalone terminal agents like Claude Code, with benchmarks from @swyx highlighting long-context reasoning as a key differentiator.

Anthropic's release of Claude Code 1.5, a terminal-native agent, is driving conversation about a fundamental shift in developer 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 consensus on agent deployment architecture is forming, with OpenAI's SDK, LangChain's protocol integrations, and Vercel's edge workers all pointing towards a standardized stack for multi-agent systems.

OpenAI, Vercel, and Replit released new SDKs and runtimes for deploying and orchestrating agents, emphasizing protocol-level standards and durable execution.

• 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 dominates the conversation, foundational model research continues with key players like MistralAI releasing high-quality open datasets critical for advancing core multimodal capabilities.

MistralAI released a large, cleaned 100M-row web OCR dataset, providing a significant public resource for training multimodal models.

• Mistral AI@MistralAIrising

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

A clear pattern is emerging where practitioners like @GregKamradt and frameworks like LlamaIndex and mem0.ai are moving beyond vector search towards structured knowledge graphs and differentiated memory stores for agents.

The discourse is evolving from simple RAG to more complex 'context engineering', addressing challenges like cache invalidation in massive context windows.

• 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 principle of durable, observable orchestration is being applied at different scales: by Temporal for dev infrastructure and by Notion/Linear for user-facing SaaS, showing a broad consolidation around workflow automation.

Workspace automation is a recurring theme, with Notion and Linear launching auto-fill and auto-triage features for productivity.

• 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 approach to prompt optimization is shifting from manual tweaking to programmatic search, with platforms like Weights & Biases providing the infrastructure to benchmark tens of thousands of variants against frontier models.

Prompt engineering is becoming more systematic, with large-scale benchmark studies from Weights & Biases and practitioner guides surfacing best practices.

• 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.

Filtering synthetic data is emerging as a critical, unsolved problem for training reliable agents, representing a more subtle challenge than standard data cleaning in supervised learning.

The focus in agent training data is on curating synthetic datasets to avoid generalization collapse, as discussed by @jerryjliu0.

• Jerry Liu@jerryjliu0repeated

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