Denoise · Twitter

The focus of agent security is escalating from simple prompt injections to complex orchestration-level vulnerabilities, as highlighted by both Anthropic and @AlexAlbert__.

Major labs like Anthropic and Google DeepMind are publicly disclosing agent jailbreaks and red-teaming frameworks for autonomous systems.

• 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.

@karpathy's observation about the shift from IDE to terminal agents is being tested in real-time, with benchmarks from @swyx and adoption reports from @levelsio.

Anthropic's release of Claude Code 1.5, a terminal-based agent, is driving conversations about its performance and a potential paradigm 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.

OpenAI's SDK, LangChain's protocol integrations, and Vercel's edge runtime show a clear convergence on standardizing the protocols and infrastructure for multi-agent systems.

A wave of new tools from OpenAI, Vercel, and Replit provides infrastructure for deploying, orchestrating, and managing AI agents as durable background workers.

• 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.

The release from a major player like Mistral AI underscores the ongoing, capital-intensive effort to build high-quality, licensed datasets, which remains a key moat in multimodal AI.

Mistral AI released a large, cleaned 100M-row web OCR dataset for public use in training models.

• Mistral AI@MistralAIrising

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

Practitioners like @GregKamradt and startups like @mem0ai argue that vector search is no longer sufficient, pushing for more structured memory systems like knowledge graphs, as demonstrated by LlamaIndex.

The conversation is shifting from basic RAG to advanced 'context engineering,' exploring cache invalidation in large contexts and new 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.

The pattern of task automation seen in Notion and Linear mirrors the goals of coding agents, suggesting a broader trend of agentic workflows being integrated into SaaS products.

Workspace automation is advancing in parallel with agent development, as Notion and Linear launch features for auto-filling and auto-triaging tasks.

• 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.

Weights & Biases is positioning prompt optimization as a formal search problem, moving beyond the anecdotal tricks shared by practitioners like @dotey to a more rigorous, data-driven approach.

Prompt engineering is maturing into a systematic practice, with scaled benchmarking of system prompts becoming a new standard for 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.

@jerryjliu0's focus on filtering synthetic data highlights a critical, non-obvious problem: ensuring that training data for agents improves, rather than poisons, their generalization capabilities.

The discussion on agent training infrastructure centers on the challenges of curating high-quality synthetic data to avoid model degradation.

• Jerry Liu@jerryjliu0repeated

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