Denoise · Twitter

Agent security has matured from theoretical prompt injection to practical pentesting and formal frameworks from labs like Anthropic and Google DeepMind.

The focus today is on red-teaming and disclosing vulnerabilities in autonomous agents, with major labs sharing frameworks and patch details.

• 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 developer workflow is fragmenting from traditional IDEs towards terminal-centric agents, a trend validated by @karpathy and early adopters like @levelsio.

Anthropic's release of Claude Code 1.5, a terminal-native agent, has dominated the conversation, with benchmarks and early adoption stories.

• 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 on agent orchestration is clear, with OpenAI, Vercel, and Replit building deployment primitives while frameworks like LangChain adapt to new protocols.

Major labs and cloud providers are shipping agent-specific SDKs, runtimes, and protocol integrations to standardize agent deployment.

• 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 agents dominate the conversation, MistralAI's dataset release indicates that building foundational resources for the open-source community remains a key priority.

A single signal today shows continued investment in foundational data, with MistralAI releasing a large-scale, open OCR dataset for training.

• Mistral AI@MistralAIrising

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

The limits of vector search at massive context scales are forcing a move toward more complex memory architectures, proposed by voices like @GregKamradt and @mem0ai.

Discussion is moving beyond simple RAG towards 'context engineering,' with new frameworks for layered memory in 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.

Beyond dedicated AI tools, established SaaS players like Notion and Linear are embedding autonomous features, suggesting a broader 'agentification' of software.

Workspace automation tools from Notion and Linear are shipping agent-like features for task management 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.

Prompt engineering is becoming more scientific, with platforms like Weights & Biases pushing the community from individual tips to data-driven optimization.

The focus in prompt engineering is shifting from anecdotal tricks to systematic, large-scale benchmarking of system prompts across 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 capabilities advance, the bottleneck is shifting to data quality, specifically filtering out synthetic data that can poison generalization, a point raised by @jerryjliu0.

The lone tweet in this category highlights the challenge of curating high-quality synthetic data for training robust agents.

• Jerry Liu@jerryjliu0repeated

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