Denoise · Twitter

The focus on agent security is shifting from theoretical exploits to formal methodologies, indicating the tech is mature enough to have a recognized attack surface.

Major labs like Anthropic and Google DeepMind are proactively releasing red-teaming frameworks and disclosures for agent 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 product category is emerging where standalone terminal agents (Claude Code) directly compete with integrated solutions (Copilot), framed by @karpathy as a fundamental workflow shift.

The release of Anthropic's Claude Code 1.5 solidifies the shift towards terminal-native agents, with benchmarks and user reports suggesting a departure from IDE-based assistants.

• 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 being built in parallel: OpenAI and Anthropic are defining protocols (like MCP), while Vercel and Replit provide the serverless runtimes to execute them.

Infrastructure providers are converging on primitives for agent orchestration and deployment, with OpenAI, Vercel, and Replit all shipping new capabilities.

• 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.

This is a quiet day focused on foundational resources rather than new models or on-device techniques, highlighting the continuous need for high-quality public data.

Activity in this category is centered on data, with Mistral AI releasing 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.

A fault line is appearing between simply using large context windows (@reach_vb) and building structured memory layers (@mem0ai, @GregKamradt), suggesting a need for more than just brute-force context.

The conversation shifts from basic RAG towards 'context engineering' and sophisticated memory architectures for agents, exploring trade-offs in large 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 pattern suggests that 'agentic' capabilities are becoming a feature, not a product, integrating autonomous task management directly into existing productivity software.

Workspace SaaS tools like Notion and Linear are shipping agent-like automation features, abstracting complex workflows into simple user-facing actions.

• 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.

Companies like Weights & Biases are treating prompt optimization as a formal ML problem, running massive experiments to find efficient frontiers, professionalizing the discipline.

The focus is on moving from anecdotal prompt 'tricks' to systematic, large-scale benchmarking of system prompts for production environments.

• 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.

This points to a second-order problem in agent development: as model capabilities grow, the bottleneck shifts to sourcing and filtering training data that doesn't poison generalization.

A single tweet highlights the nuanced problem of curating high-quality synthetic data specifically for training autonomous agents.

• Jerry Liu@jerryjliu0repeated

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