Denoise · Twitter

Major labs like AnthropicAI and GoogleDeepMind are establishing security best practices for agents in parallel with their release, indicating the maturity of the ecosystem.

Red teaming frameworks and responsible disclosures are being released for autonomous agents, focusing on prompt injection, tool leakage, and orchestration-layer 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.

There is a clear convergence between high-level commentary from figures like karpathy and concrete product releases like Anthropic's, with early adopters like levelsio already reporting productivity gains.

Anthropic's release of Claude Code 1.5, a terminal-native agent, anchors a broader discussion on the shift away from traditional IDEs for development 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.

The ecosystem is converging on a multi-worker, orchestrated agent model, with LangChain providing protocol compatibility layers between different platforms like Anthropic's MCP.

Major infrastructure providers including OpenAI, Vercel, and Replit are releasing SDKs and runtimes to standardize the deployment and orchestration 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 most of the day's focus is on agent tooling, MistralAI's contribution is a foundational data asset, essential for building future multimodal capabilities.

MistralAI released a large-scale, cleaned 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.

A new vocabulary is emerging with 'context engineering' (GregKamradt) and distinct working vs. long-term memory (mem0ai), signaling a move beyond simple vector search RAG.

The discussion is shifting from raw context length to sophisticated memory architectures, exploring cache invalidation, retrieval strategies, and layered memory systems.

• 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 trend towards automation is visible in both user-facing SaaS (Notion, Linear) and backend infrastructure (Temporal), reflecting a broad demand for autonomous, resilient processes.

Workspace automation features are becoming standard in SaaS tools like Notion and Linear, while Temporal highlights durable execution for complex, stateful workflows.

• 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.

A split is visible between the craft of prompt writing shared by practitioners like dotey and the industrial approach taken by platforms like Weights & Biases.

The practice of prompt engineering is maturing from individual tricks to large-scale, systematic benchmarking of system prompts against multiple models.

• 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.

As agent reasoning capabilities advance, the bottleneck is shifting back to the quality and filtering of training data, a classic machine learning problem highlighted by jerryjliu0.

The focus in agent training is on high-quality data curation to avoid generalization failure from poisoned synthetic data.

• Jerry Liu@jerryjliu0repeated

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