Denoise · Twitter

The security focus is shifting from prompt injection within the model to vulnerabilities in the agent's orchestration and tool-use layer, a concern shared by @GoogleDeepMind and @AnthropicAI.

Major labs and independent researchers are publishing formal frameworks and disclosures for securing autonomous agents against new attack vectors.

• 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 consensus is forming around the terminal as the primary interface for coding agents, with @karpathy articulating the paradigm shift that @AnthropicAI's product embodies.

Anthropic's launch of Claude Code 1.5, a terminal-native agent, marks a major shift in developer tooling, validated by influential figures and early adopters.

• 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 complete, de facto agent stack is rapidly converging, with @OpenAI providing orchestration primitives and @Vercel/@replit offering serverless deployment targets for agent workers.

A wave of new releases from OpenAI, Vercel, and Replit provides the SDKs, runtimes, and deployment harnesses for the new class 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 agent orchestration dominates the conversation, foundational model providers like @MistralAI continue to invest in core data assets that enable multimodal capabilities.

Mistral AI contributed to the open-source community by releasing a large-scale, cleaned dataset for web OCR training.

• Mistral AI@MistralAIrising

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

The limitations of basic vector retrieval are becoming a key bottleneck, pushing projects like @mem0ai and commentary from @GregKamradt toward structured, multi-layered memory systems.

The discourse is moving past simple RAG, with experts proposing more sophisticated 'context engineering' frameworks and new memory architectures for 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 concept of durable, replayable workflows from infrastructure tools like @temporalio is converging with the practical needs of business automation seen in @NotionHQ and @linear.

Workspace automation tools like Notion and Linear are launching features for chained, autonomous updates, mirroring agent orchestration patterns.

• 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 approach is bifurcating: tactical, human-centric tips from creators like @dotey exist alongside large-scale, automated benchmark studies from platforms like @weights_biases.

System prompt engineering is maturing from a collection of tricks to a systematic, large-scale benchmarking and optimization problem.

• 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 improves, the bottleneck returns to data quality, with @jerryjliu0 pointing out that poor synthetic data can actively poison a model's ability to generalize.

A practitioner highlights the critical but difficult task of curating synthetic datasets for agent training to avoid performance degradation.

• Jerry Liu@jerryjliu0repeated

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