Denoise · Twitter

The field is converging on agent-specific attack surfaces, moving beyond simple prompt injection to complex orchestration-layer exploits, as noted by @AnthropicAI and @GoogleDeepMind.

Discussion centers on red-teaming autonomous agents, with major labs releasing frameworks and public disclosures on 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.

Anthropic's Claude Code 1.5, combined with validation from @karpathy and @swyx, signals a convergence on the terminal as the new primary interface for AI-assisted development.

Major releases focus on terminal-native coding agents that integrate directly into developer workflows, challenging traditional IDEs.

• 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 ecosystem is converging on a standard agent stack: protocol-level tool use (OpenAI), orchestration frameworks (LangChain), and managed deployment (Vercel/Replit).

New SDKs and deployment platforms from OpenAI, Vercel, and Replit are being released to support the orchestration and hosting of 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 agent infrastructure dominates today's conversation, @MistralAI's contribution shows continued investment in foundational datasets, a critical dependency for future multimodal agents.

A single major data release from MistralAI provides a large-scale, open dataset for web OCR training.

• Mistral AI@MistralAIrising

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

Actors like @GregKamradt and @mem0ai are fragmenting the monolithic 'RAG' concept into a toolbox of specific techniques like caching, graph retrieval, and differentiated memory stores.

The conversation is evolving from simple RAG implementations to more complex 'context engineering' and specialized agent memory architectures.

• 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.

While AI-native companies build explicit agent frameworks, established SaaS players like @NotionHQ and @linear are converging on similar autonomous capabilities from a workflow automation angle.

Workspace automation tools like Notion and Linear are shipping agent-like features for task management and triage.

• 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 discipline is maturing from individual craft (@dotey) to industrial-scale engineering, with platforms like @weights_biases enabling rigorous, data-driven prompt optimization.

The focus is shifting from anecdotal prompt tricks to systematic, large-scale benchmarking to find optimal system prompts.

• 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.

@jerryjliu0's post points to a critical bottleneck: as model capabilities advance, the core problem shifts from raw compute to the nuanced task of data curation.

The key challenge highlighted is the curation of high-quality training data for agents, particularly filtering harmful synthetic data.

• Jerry Liu@jerryjliu0repeated

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