Denoise · Twitter

The conversation on agent security has shifted from theoretical risks to practical, operational disciplines, with both offensive frameworks and defensive disclosures becoming standard practice.

Major labs like Anthropic and DeepMind are formalizing agent security, releasing red-teaming frameworks and public write-ups on patched jailbreaks.

• 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.

The primary developer interface is being contested, with Anthropic's terminal agent and commentary from @karpathy suggesting a move toward a conversational, command-line-centric workflow.

Anthropic's launch of the terminal-native Claude Code 1.5 agent is driving discussion about a fundamental shift in developer workflows away from traditional IDEs.

• 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 convergence pattern is emerging as OpenAI, Anthropic (via MCP), and cloud platforms like Vercel race to provide the foundational orchestration layer for multi-agent systems.

Major infrastructure providers including OpenAI, Vercel, and Replit are releasing standardized SDKs and deployment runtimes for building and orchestrating 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 the category is quiet today, MistralAI's data release provides foundational material for training next-generation vision models, continuing their strategy of enabling the open-source community.

MistralAI released a large-scale, 100M-row web OCR dataset for public use, providing a new resource for training vision models.

• Mistral AI@MistralAIrising

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

Practitioners like @GregKamradt and @reach_vb are finding that massive context windows aren't a silver bullet, forcing a shift towards more sophisticated memory and retrieval strategies.

The discussion is moving past simple RAG, focusing on 'context engineering' and advanced memory architectures to handle multi-million token 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 patterns of agentic automation are being integrated into established SaaS products like Notion and Linear, normalizing workflow automation for non-technical users.

Mainstream productivity tools like Notion and Linear are embedding agent-like automation features for tasks like data entry 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 practice of prompt engineering is becoming more data-driven, with players like Weights & Biases applying large-scale experimentation to find optimal system prompts.

Efforts are focused on systematically improving and benchmarking system prompts at scale, with both practical tips and large-scale study results being shared.

• 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.

The focus on dataset quality, articulated by @jerryjliu0, indicates a growing maturity in agent development, moving from architectural novelty to data-centric problems.

A specific challenge in training effective agents is highlighted: curating high-quality synthetic data and filtering out examples that harm generalization.

• Jerry Liu@jerryjliu0repeated

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