Denoise · Twitter

A clear convergence is visible as major labs like @AnthropicAI and @GoogleDeepMind are now publishing formal red-teaming frameworks and disclosures for agents, establishing industry norms.

The focus is squarely on defining and mitigating the new attack surfaces introduced by autonomous agents.

• 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.

@AnthropicAI's release, validated by adoption from @levelsio and benchmarks from @swyx, creates a new fault-line against the established IDE-plugin model of Copilot and Cursor.

The launch of Anthropic's terminal-native Claude Code agent signals a major shift in the developer tool landscape.

• 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 convergence pattern is clear: @OpenAI defines the protocol, @LangChainAI provides the integration framework, and platforms like @vercel and @replit offer the deployment runtime.

The ecosystem is rapidly building the infrastructure layer for deploying and orchestrating agents, following protocol releases from major labs.

• 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 release from @MistralAI is a classic ecosystem-enabling move, providing a foundational asset that allows smaller players to compete with the data moats of larger, closed-source labs.

Mistral AI released a large-scale, open OCR dataset for training vision models.

• Mistral AI@MistralAIrising

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

A fault-line is emerging between scaling raw context (@reach_vb) and architecting smarter memory systems (@GregKamradt, @mem0ai), with graph-based retrieval (@llamaindex) as a third path.

The discussion is evolving from simply increasing context window size to designing more sophisticated context engineering and 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.

A convergence is happening where SaaS tools like @NotionHQ and @linear are building workflow automation that mirrors the durable execution patterns championed by specialized orchestration tools like @temporalio.

Workspace automation tools are independently converging on agent-like capabilities for task management and data processing.

• 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 clear progression is visible from individual prompt-writing advice (@dotey) to industrial-scale system prompt optimization and analysis from platforms like @weights_biases.

The practice of prompt engineering is maturing towards systematic, large-scale benchmarking over anecdotal tricks.

• 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 analysis highlights a key bottleneck in the agent development pipeline: filtering synthetic data to avoid generalization poisoning is becoming a crucial, specialized skill.

The focus in agent training is shifting towards the critical importance of high-quality, curated datasets over raw data volume.

• Jerry Liu@jerryjliu0repeated

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