Denoise · Twitter

The security conversation is shifting from abstract prompt injection to concrete vulnerabilities in agent orchestration layers, as detailed by @AnthropicAI and @MalwareTechBlog.

The focus is on agent security, with major labs like Anthropic and GoogleDeepMind releasing detailed write-ups on red-teaming and real-world 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 convergence toward the terminal as the new IDE is clear, with @karpathy providing the theory and @AnthropicAI shipping the product, directly challenging Cursor and Copilot.

This category is dominated by the launch of Anthropic's Claude Code 1.5, sparking widespread discussion about the superiority of terminal-native agent 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.

A new standardized stack is emerging, with @OpenAI and @AnthropicAI defining protocols that frameworks like @LangChainAI integrate, and platforms like @vercel and @replit providing the hosting.

Major infrastructure players are shipping tools for agent orchestration and deployment, including SDKs, edge runtimes, and deployment harnesses.

• 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 conversation is on agentic reasoning, @MistralAI's OCR dataset release shows continued investment in foundational perception capabilities, a critical but less hyped area.

A quiet day with a single, significant data release from MistralAI for web-scale optical character recognition.

• Mistral AI@MistralAIrising

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

Practitioners like @GregKamradt and tools like @mem0ai are fragmenting the monolithic RAG concept, exploring specialized architectures for caching, retrieval, and memory differentiation.

The discourse is evolving from simple RAG implementations to a more sophisticated discipline of 'context engineering' for large 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.

While AI agents capture the narrative, established tools like @NotionHQ and @linear continue to ship practical, deterministic automation, indicating a parallel path to user value.

Workspace automation sees new releases from major SaaS players like Notion and Linear, focusing on non-LLM workflow and triage features.

• 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 push for rigor, exemplified by @weights_biases running 40k variants, suggests a convergence on empirical validation over anecdotal 'prompt tricks'.

The focus is on moving prompt engineering from an art to a science through large-scale, systematic benchmarking.

• 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 capabilities advance, the bottleneck shifts back to data quality. @jerryjliu0's point on synthetic data poisoning reveals a key challenge for underlying ML infrastructure.

A single tweet highlights the critical, often overlooked problem of curating high-quality datasets for training agents.

• Jerry Liu@jerryjliu0repeated

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