Denoise · Twitter

The focus on agent security from @AnthropicAI and @GoogleDeepMind signals that safety is rapidly becoming a key competitive differentiator, not just a compliance issue.

Major labs are proactively disclosing agent vulnerabilities and publishing red-teaming frameworks for this new class of software.

• 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 launch establishes a direct challenge to the Copilot/IDE paradigm, signaling a convergence on owning the developer's entire inner loop.

The market is shifting from simple code completion to full-fledged, terminal-native coding agents with file system access and reasoning capabilities.

• 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 clear pattern of convergence is visible between @OpenAI, @LangChainAI, @vercel, and @replit, all building tooling for a new 'durable agent worker' abstraction.

Infrastructure providers are racing to ship agent-native deployment harnesses, orchestration primitives, and specialized runtimes.

• 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 a significant data release, the focus from @MistralAI on datasets contrasts with the broader industry's heavy push into agentic systems and tooling today.

Mistral AI released a large-scale open dataset for web OCR, providing a key resource for multimodal model training.

• Mistral AI@MistralAIrising

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

The ideas from @GregKamradt and @mem0ai suggest a divergence from simple vector search, with more complex systems like knowledge graphs (@llamaindex) becoming critical for agent memory.

The conversation is evolving from simple RAG towards sophisticated 'context engineering' and complex memory architectures for stateful agents.

• 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 quiet launch of auto-triage in @linear and automation in @NotionHQ shows how agentic principles are being embedded into SaaS workflows, not just developer tools.

Workspace automation tools like Notion and Linear are embedding agent-like capabilities for auto-filling and issue 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 work detailed by @weights_biases represents a move from artisanal prompt crafting to an industrial, data-driven optimization process, treating prompts as hyperparameters.

The focus in prompt engineering is shifting towards systematic, large-scale benchmarking of system prompts to find optimal performance.

• 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 tweet serves as a reminder that behind the agent tooling race, fundamental data quality problems remain a critical bottleneck for model generalization.

A niche but important discussion highlights the challenges of dataset curation for agent training, specifically filtering harmful synthetic data.

• Jerry Liu@jerryjliu0repeated

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