diff --git a/README.md b/README.md
index 3fce7f9..da9d6ee 100644
--- a/README.md
+++ b/README.md
@@ -63,6 +63,8 @@ _Opinionated. No tutorials, no listicles, no marketing. Continuously maintained
 - [smolagents](https://huggingface.co/docs/smolagents/index) — Hugging Face. Minimalist code-acting agent library.
 - [Mastra](https://mastra.ai/docs) — TypeScript-first.
 - [Inngest AgentKit](https://agentkit.inngest.com/) — TS framework on top of Inngest's durable runtime.
+- [How to Use RLMs in Deep Agents](https://www.langchain.com/blog/how-to-use-rlms-in-deep-agents) — LangChain. Recursive language models (RLMs) address context rot by having agents write code that fans out subagents over context chunks — grep/map/reduce patterns over large inputs. Benchmarked on OOLONG long-context reasoning; shows where turn-by-turn agents break down and dynamic subagent dispatch holds up.
+- [Building LangGraph: Designing an Agent Runtime from First Principles](https://www.langchain.com/blog/building-langgraph) — LangChain. First-principles design rationale behind LangGraph: why graph-based state machines, how control flow and durability are handled, and what tradeoffs shaped the runtime's architecture for production agents.
 
 ## 4. Durable execution for agents
 
@@ -73,6 +75,8 @@ _Opinionated. No tutorials, no listicles, no marketing. Continuously maintained
 - [Restate — AI agents](https://docs.restate.dev/use-cases/ai-agents) — Lightweight durable execution; agents as virtual objects.
 - [Hatchet — Durable Tasks](https://docs.hatchet.run/v1/durable-tasks) — Postgres-backed task queue with agent-aware patterns (agentic loops, HITL).
 - [DBOS — Durable Execution for Building Crashproof AI Agents](https://www.dbos.dev/blog/durable-execution-crashproof-ai-agents) — Postgres-as-runtime; smaller-team alternative to Temporal.
+- [How we built saga rollbacks for Cloudflare Workflows](https://blog.cloudflare.com/rollbacks-for-workflows/) — Cloudflare. Saga-pattern rollbacks in Cloudflare Workflows: each `step.do()` accepts a compensating action, executed in reverse order on failure. Covers implementation trade-offs, durability guarantees, and how compensation state is checkpointed alongside forward progress.
+- [Fault Tolerance in LangGraph: Retries, Timeouts and Error Handlers](https://www.langchain.com/blog/fault-tolerance-in-langgraph) — LangChain. Three fault-tolerance primitives inside LangGraph — `RetryPolicy` (backoff retries), `TimeoutPolicy` (wall-clock and idle caps), and `error_handler` (post-retry cleanup) — and how they compose. Applies the SAGA pattern to multi-step workflows with real-world side effects.
 
 ## 5. Memory systems
 
@@ -83,6 +87,7 @@ _Opinionated. No tutorials, no listicles, no marketing. Continuously maintained
 - [getzep/graphiti](https://github.com/getzep/graphiti) — The temporal-graph engine standalone.
 - [LangMem](https://langchain-ai.github.io/langmem/) — Semantic / episodic / procedural primitives over LangGraph stores.
 - [Generative Agents (Park et al.)](https://arxiv.org/abs/2304.03442) — Reflection + episodic memory; still the best single read.
+- [Context Management for Deep Agents](https://www.langchain.com/blog/context-management-for-deepagents) — LangChain. How Deep Agents SDK handles context rot in long-running tasks: offloading, summarization, and filesystem abstraction as first-class context management primitives.
 
 ## 6. Sandboxing & code execution
 
@@ -95,6 +100,8 @@ _Opinionated. No tutorials, no listicles, no marketing. Continuously maintained
 - [hyperlight-dev/hyperlight](https://github.com/hyperlight-dev/hyperlight) — Microsoft's sub-millisecond WASM/VM micro-sandbox.
 - [gVisor docs](https://gvisor.dev/docs/) — User-space kernel; understand it before trusting "sandboxed" claims.
 - [Interpreters in Deep Agents: Code Between Tool Calls and Sandboxes](https://www.langchain.com/blog/give-your-agents-an-interpreter) — LangChain. Embedded interpreter runtimes let agents write code to coordinate tool calls, manage working state between steps, and control what gets surfaced into model context — reducing token pressure and enabling finer-grained orchestration than pure tool-dispatch.
+- [How Deep Agents Run Untrusted Code Without a Sandbox](https://www.langchain.com/blog/running-untrusted-agent-code-without-a-sandbox) — LangChain. In-process isolation via WASM + QuickJS as a lighter alternative to full VM/container sandboxes — covers least-privilege capability gating and snapshot-based durable pauses for long-running agent tasks.
+- [How to Choose the Right Sandbox for AI Agents](https://www.langchain.com/blog/how-to-choose-the-right-sandbox-for-your-agent) — LangChain. Decision framework for sandbox selection: filesystem isolation, network access controls, resource limits, and microVM trade-offs for code-executing agents.
 
 ## 7. Inference & gateway infrastructure
 
@@ -118,6 +125,13 @@ _Opinionated. No tutorials, no listicles, no marketing. Continuously maintained
 - [Who Validates the Validators? (EvalGen)](https://arxiv.org/abs/2404.12272) — Shankar et al. Critical paper on grader drift.
 - [Judging LLM-as-a-Judge (MT-Bench)](https://arxiv.org/abs/2306.05685) — The original position/verbosity/self-preference bias paper.
 - [Low-Hanging Fruit for RAG Search](https://jxnl.co/writing/2024/05/11/low-hanging-fruit-for-rag-search/) — Jason Liu. Retrieval-side instrumentation.
+- ["It's Hard to Eval" Is a Product Smell](https://hamel.dev/blog/posts/eval-smell/) — Hamel Husain. Difficulty eval-ing is a design flaw: artifacts hard to verify by machine are hard to verify by users too. Three worked examples (data agent, PE curriculum builder, workers-comp report) show how designing for human verifiability — provenance, progressive disclosure, diff-scoped review — also makes automated evals tractable.
+- [Designing Efficient Verifiers for Legal Agents](https://www.langchain.com/blog/designing-efficient-verifiers-for-legal-agents) — Harvey & LangChain Labs. Harvey + LangChain Labs study on building cheaper, more reliable LLM-as-judge verifiers for legal-agent eval and post-training signal — covering verifier calibration, cost-accuracy tradeoffs, and failure modes in domain-specific agentic pipelines.
+- [How we build evals for Deep Agents](https://www.langchain.com/blog/how-we-build-evals-for-deep-agents) — LangChain. Data curation, metrics selection, and testing strategies for evaluating deep/long-horizon agents — targeted behavioral measurement over surface-level accuracy.
+- [Building a 100x Cheaper Trace Judge with Fireworks](https://www.langchain.com/blog/building-a-100x-cheaper-trace-judge-with-fireworks) — LangChain. Fine-tune a small open model on production trace signals to match frontier judge performance at ~1/100th the cost — covers data mining from traces, distillation pipeline, and eval results.
+- [Iterating Towards LLM Reliability with Evaluation Driven Development](https://www.langchain.com/blog/iterating-towards-llm-reliability-with-evaluation-driven-development) — LangChain / Dosu. Evaluation-driven development (EDD) as a production discipline: build eval suites before shipping, use production traces to seed regression datasets, and gate iterative improvements on measurable score deltas. Practical case study from Dosu at scale.
+- [Aligning LLM-as-a-Judge with Human Preferences](https://www.langchain.com/blog/aligning-llm-as-a-judge-with-human-preferences) — LangChain. Few-shot calibration techniques for closing the gap between LLM-judge scores and human preference labels — with concrete implementation in LangSmith's self-improving evaluators.
+- [Evaluating Deep Agents: Our Learnings](https://www.langchain.com/blog/evaluating-deep-agents-our-learnings) — LangChain. Five evaluation patterns for long-horizon agents: bespoke unit tests, single-step tool validation, full-turn checks, multi-turn simulations, and environment setup — grounded in production experience.
 
 ## 9. Evaluation — frameworks & benchmarks
 
@@ -133,6 +147,9 @@ _Opinionated. No tutorials, no listicles, no marketing. Continuously maintained
 - [Langfuse Evaluations](https://langfuse.com/docs/evaluation/overview) — **OS + SaaS.**
 - [Patronus AI](https://docs.patronus.ai:443/docs) — **SaaS.** Managed judge models (Lynx for hallucination).
 - **Benchmarks:** [SWE-bench](https://www.swebench.com/), [SWE-bench Verified](https://openai.com/index/introducing-swe-bench-verified/), [GAIA](https://arxiv.org/abs/2311.12983), [τ-bench](https://github.com/sierra-research/tau-bench), [WebArena](https://webarena.dev/), [OSWorld](https://os-world.github.io/), [MLE-bench](https://github.com/openai/mle-bench), [SWE-Lancer](https://arxiv.org/abs/2502.12115).
+- [Patterns for Building Cybersecurity Evals](https://eugeneyan.com//writing/cybersecurity-evals/) — Eugene Yan. Four-component pattern for adversarial evals: sandboxed target, difficulty-modulating inputs, tools, and a grader. Concrete design decisions for evaluating agents on cybersecurity tasks.
+- [Benchmarking Single Agent Performance](https://www.langchain.com/blog/react-agent-benchmarking) — LangChain. Empirical study of how scaling instruction count and tool count degrades ReAct agent performance across claude-3.5-sonnet, gpt-4o, o1, and o3-mini — two task domains, reproducible methodology.
+- [Evaluating Large Language Models With OpenEvals](https://www.langchain.com/blog/evaluating-llms-with-openevals) — LangChain. Pre-built evaluators for LLM-as-judge, structured output correctness, and agent trajectory assessment via OpenEvals and AgentEvals; covers scorer design, integration patterns, and production deployment.
 
 ## 10. Observability & tracing
 
@@ -164,6 +181,7 @@ _Opinionated. No tutorials, no listicles, no marketing. Continuously maintained
 - [NIST AI RMF + Generative AI Profile](https://www.nist.gov/itl/ai-risk-management-framework) — Risk-management vocabulary auditors will use.
 - [MITRE ATLAS](https://atlas.mitre.org/) — ATT&CK-style matrix for ML/agent threats.
 - [Trail of Bits — Prompt injection to RCE in AI agents](https://blog.trailofbits.com/2025/10/22/prompt-injection-to-rce-in-ai-agents/) — Recent, concrete RCE chain.
+- [Computer-Use and TOCTOU: What You Click Is Not What You Get!](https://embracethered.com/blog/posts/2026/toctou-agent-what-you-click-is-not-what-you-get/) — embrace-the-red (Johann Rehberger). TOCTOU race condition attack against computer-use agents: content shown to the user at approval time can be swapped before the agent acts on it, breaking human-in-the-loop trust assumptions. Reproduces the ChatGPT Operator vulnerability disclosed by Jun Kokatsu with a full attack chain demo.
 
 ## 13. Coding agent infrastructure (read for harness design even if not building one)
 
@@ -178,6 +196,8 @@ _Opinionated. No tutorials, no listicles, no marketing. Continuously maintained
 - [Geoffrey Huntley — how to build a coding agent (workshop)](https://ghuntley.com/agent/) — Free workshop on building one from scratch.
 - [Geoffrey Huntley — Ralph Wiggum loop](https://ghuntley.com/ralph/) — The brute-force feedback-loop pattern essay.
 - [Open SWE: An Open-Source Framework for Internal Coding Agents](https://www.langchain.com/blog/open-swe-an-open-source-framework-for-internal-coding-agents) — LangChain. Open-source SWE-agent framework built on LangGraph; covers core architectural components — task manager, programmer agent, and sandboxed execution — for deploying internal coding agents at scale.
+- [Build your own vulnerability harness](https://blog.cloudflare.com/build-your-own-vulnerability-harness/) — Cloudflare. Multi-stage vulnerability discovery harness: state management, adversarial review to suppress false positives, and strategies for routing around LLM context limits in an automated triage loop.
+- [How to turn Claude Code into a domain specific coding agent](https://www.langchain.com/blog/how-to-turn-claude-code-into-a-domain-specific-coding-agent) — LangChain. Context engineering techniques for specializing Claude Code toward domain-specific libraries — CLAUDE.md, tool allowlists, prompt injection — with eval results comparing approaches.
 
 ## 14. SRE & operations agents (K8s, observability, IaC)
 
diff --git a/resources.yaml b/resources.yaml
index 88fdeec..582daaf 100644
--- a/resources.yaml
+++ b/resources.yaml
@@ -96,7 +96,8 @@ sections:
       and RCE write-ups for agents.
   - id: coding-agent-infra
     order: 13
-    title: "Coding agent infrastructure (read for harness design even if not building one)"
+    title: "Coding agent infrastructure (read for harness design even if not building
+      one)"
     description: |
       Harness design, hook systems, tool allowlists, and operational patterns
       for building and running coding agents — Claude Code best practices, the
@@ -123,19 +124,23 @@ top_7:
 worth_following:
   - name: Anthropic Engineering
     url: https://www.anthropic.com/engineering
-    blurb: Engineering posts from the model-maker — agents, tooling, and operational patterns.
+    blurb: Engineering posts from the model-maker — agents, tooling, and 
+      operational patterns.
     verified_at: &id001 2026-06-29
   - name: Cognition blog
     url: https://cognition.com/blog
-    blurb: Case studies on building Devin; the productive disagreement to "Don't Build Multi-Agents."
+    blurb: Case studies on building Devin; the productive disagreement to "Don't
+      Build Multi-Agents."
     verified_at: *id001
   - name: Simon Willison
     url: https://simonwillison.net/
-    blurb: The field's running curator — daily-ish takes on LLMs, agents, and prompt injection.
+    blurb: The field's running curator — daily-ish takes on LLMs, agents, and 
+      prompt injection.
     verified_at: *id001
   - name: Embrace The Red
     url: https://embracethered.com/blog/
-    blurb: Johann Rehberger's red-team blog — concrete agent exploits and bypass techniques.
+    blurb: Johann Rehberger's red-team blog — concrete agent exploits and bypass
+      techniques.
     verified_at: *id001
   - name: Cloudflare AI agents tag
     url: https://blog.cloudflare.com/tag/ai-agents/
@@ -173,11 +178,13 @@ resources:
     superseded_by: null
     notes: null
     top_7_author: "Anthropic (Schluntz/Zhang)"
-    top_7_blurb: 'The "workflows vs agents" mental model that everything else builds on.'
+    top_7_blurb: 'The "workflows vs agents" mental model that everything else builds
+      on.'
 
   - id: effective-context-engineering
     section: foundations
-    url: https://www.anthropic.com/engineering/effective-context-engineering-for-ai-agents
+    url: 
+      https://www.anthropic.com/engineering/effective-context-engineering-for-ai-agents
     title: "Effective context engineering for AI agents"
     author: "Anthropic"
     type: article
@@ -191,7 +198,8 @@ resources:
     paywall: false
     superseded_by: null
     notes: null
-    top_7_blurb: "The successor concept to prompt engineering; defines the actual job."
+    top_7_blurb: "The successor concept to prompt engineering; defines the actual
+      job."
 
   - id: writing-effective-tools
     section: foundations
@@ -200,7 +208,8 @@ resources:
     author: "Anthropic"
     type: article
     license: null
-    blurb: "Tools as contracts between deterministic and non-deterministic systems; eval-driven tool refinement."
+    blurb: "Tools as contracts between deterministic and non-deterministic systems;
+      eval-driven tool refinement."
     cluster: null
     tags: [tools]
     added_at: 2026-05-03
@@ -212,12 +221,14 @@ resources:
 
   - id: effective-harnesses
     section: foundations
-    url: https://www.anthropic.com/engineering/effective-harnesses-for-long-running-agents
+    url: 
+      https://www.anthropic.com/engineering/effective-harnesses-for-long-running-agents
     title: "Effective harnesses for long-running agents"
     author: "Anthropic"
     type: article
     license: null
-    blurb: "Initializer + worker pattern for cross-context continuity (claude-progress.txt + git)."
+    blurb: "Initializer + worker pattern for cross-context continuity (claude-progress.txt
+      + git)."
     cluster: null
     tags: [harness, long-running]
     added_at: 2026-05-03
@@ -263,7 +274,8 @@ resources:
 
   - id: openai-practical-guide
     section: foundations
-    url: https://cdn.openai.com/business-guides-and-resources/a-practical-guide-to-building-agents.pdf
+    url: 
+      https://cdn.openai.com/business-guides-and-resources/a-practical-guide-to-building-agents.pdf
     title: "A practical guide to building agents (PDF)"
     author: "OpenAI"
     type: article
@@ -296,7 +308,8 @@ resources:
     superseded_by: null
     notes: null
     top_7_author: "Dex Horthy / HumanLayer"
-    top_7_blurb: 'Heroku''s 12-factor reframed for LLM systems; the canonical "agents are mostly software" doctrine.'
+    top_7_blurb: 'Heroku''s 12-factor reframed for LLM systems; the canonical "agents
+      are mostly software" doctrine.'
 
   - id: twelve-factor-agents-talk
     section: foundations
@@ -331,7 +344,8 @@ resources:
     paywall: false
     superseded_by: null
     notes: null
-    top_7_blurb: "Read alongside #4 — the productive disagreement at the heart of agent architecture in 2025/26."
+    top_7_blurb: "Read alongside #4 — the productive disagreement at the heart of
+      agent architecture in 2025/26."
 
   - id: agentic-engineering-patterns
     section: foundations
@@ -357,7 +371,8 @@ resources:
     author: "Simon Willison"
     type: article
     license: null
-    blurb: "Private data + untrusted content + exfiltration channel = the agent threat model."
+    blurb: "Private data + untrusted content + exfiltration channel = the agent threat
+      model."
     cluster: null
     tags: [security, threat-model]
     added_at: 2026-05-03
@@ -457,7 +472,8 @@ resources:
 
   - id: mcp-authorization
     section: tools-mcp
-    url: https://modelcontextprotocol.io/specification/2025-11-25/basic/authorization
+    url: 
+      https://modelcontextprotocol.io/specification/2025-11-25/basic/authorization
     title: "MCP — Authorization"
     author: null
     type: spec
@@ -559,7 +575,8 @@ resources:
 
   - id: a2a-announcement
     section: tools-mcp
-    url: https://developers.googleblog.com/en/a2a-a-new-era-of-agent-interoperability/
+    url: 
+      https://developers.googleblog.com/en/a2a-a-new-era-of-agent-interoperability/
     title: "Announcing the Agent2Agent Protocol (A2A)"
     author: "Google"
     type: article
@@ -608,7 +625,9 @@ resources:
     archived: false
     paywall: false
     superseded_by: null
-    notes: "Permanent home is eval-philosophy; kept hidden so the rendered output round-trips cleanly against the original (original has no section listing for this entry)."
+    notes: "Permanent home is eval-philosophy; kept hidden so the rendered output
+      round-trips cleanly against the original (original has no section listing for
+      this entry)."
     hidden: true
 
   # ----------------------------------------------------------------------
@@ -735,7 +754,8 @@ resources:
 
   - id: autogen-v04
     section: multi-agent-frameworks
-    url: https://microsoft.github.io/autogen/stable/user-guide/core-user-guide/index.html
+    url: 
+      https://microsoft.github.io/autogen/stable/user-guide/core-user-guide/index.html
     title: "AutoGen v0.4"
     author: null
     type: docs
@@ -891,7 +911,8 @@ resources:
 
   - id: temporal-durable-execution-ai
     section: durable-execution
-    url: https://temporal.io/blog/durable-execution-meets-ai-why-temporal-is-the-perfect-foundation-for-ai
+    url: 
+      https://temporal.io/blog/durable-execution-meets-ai-why-temporal-is-the-perfect-foundation-for-ai
     title: "Temporal — Durable Execution meets AI"
     author: null
     type: article
@@ -925,7 +946,8 @@ resources:
 
   - id: inngest-durable-execution
     section: durable-execution
-    url: https://www.inngest.com/blog/durable-execution-key-to-harnessing-ai-agents
+    url: 
+      https://www.inngest.com/blog/durable-execution-key-to-harnessing-ai-agents
     title: "Inngest — Durable Execution: The Key to Harnessing AI Agents in Production"
     author: null
     type: article
@@ -996,7 +1018,8 @@ resources:
   # ----------------------------------------------------------------------
   - id: anthropic-memory-tool
     section: memory-systems
-    url: https://platform.claude.com/docs/en/build-with-claude/tool-use/memory-tool
+    url: 
+      https://platform.claude.com/docs/en/build-with-claude/tool-use/memory-tool
     title: "Anthropic — Memory tool"
     author: null
     type: docs
@@ -1333,7 +1356,8 @@ resources:
     author: null
     type: docs
     license: null
-    blurb: "100+ provider proxy; OpenAI-shaped API; the boring-but-essential routing layer."
+    blurb: "100+ provider proxy; OpenAI-shaped API; the boring-but-essential routing
+      layer."
     cluster: null
     tags: [gateway, routing]
     added_at: 2026-05-03
@@ -1577,7 +1601,8 @@ resources:
     author: "UK AISI"
     type: docs
     license: "OS"
-    blurb: "**OS.** Best-in-class for agent evals; sandboxed tool use, MCP support, used by Anthropic/DeepMind."
+    blurb: "**OS.** Best-in-class for agent evals; sandboxed tool use, MCP support,
+      used by Anthropic/DeepMind."
     cluster: null
     tags: [evals, inspect-ai]
     added_at: 2026-05-03
@@ -1905,7 +1930,8 @@ resources:
     author: null
     type: spec
     license: null
-    blurb: "**Foundational.** The vendor-neutral schema for LLM/agent spans. Build to this and swap backends."
+    blurb: "**Foundational.** The vendor-neutral schema for LLM/agent spans. Build
+      to this and swap backends."
     cluster: null
     tags: [observability, opentelemetry]
     added_at: 2026-05-03
@@ -2110,7 +2136,8 @@ resources:
 
   - id: langfuse-cost-tracking
     section: production-testing
-    url: https://langfuse.com/docs/observability/features/token-and-cost-tracking
+    url: 
+      https://langfuse.com/docs/observability/features/token-and-cost-tracking
     title: "Langfuse — Cost tracking"
     author: null
     type: docs
@@ -2284,7 +2311,8 @@ resources:
 
   - id: trailofbits-prompt-injection-rce
     section: security
-    url: https://blog.trailofbits.com/2025/10/22/prompt-injection-to-rce-in-ai-agents/
+    url: 
+      https://blog.trailofbits.com/2025/10/22/prompt-injection-to-rce-in-ai-agents/
     title: "Trail of Bits — Prompt injection to RCE in AI agents"
     author: null
     type: article
@@ -2326,7 +2354,8 @@ resources:
     author: "Anthropic"
     type: docs
     license: null
-    blurb: "Hooks (PreToolUse/PostToolUse/Stop/etc.), tool allowlists, custom tools as in-process MCP."
+    blurb: "Hooks (PreToolUse/PostToolUse/Stop/etc.), tool allowlists, custom tools
+      as in-process MCP."
     cluster: null
     tags: [claude-agent-sdk, harness]
     added_at: 2026-05-03
@@ -2589,12 +2618,15 @@ resources:
     top_7_blurb: "The best single multi-agent case study, with concrete failure modes."
   - id: open-swe-langchain
     section: coding-agent-infra
-    url: https://www.langchain.com/blog/open-swe-an-open-source-framework-for-internal-coding-agents
+    url: 
+      https://www.langchain.com/blog/open-swe-an-open-source-framework-for-internal-coding-agents
     title: 'Open SWE: An Open-Source Framework for Internal Coding Agents'
     author: LangChain
     type: article
     license: null
-    blurb: Open-source SWE-agent framework built on LangGraph; covers core architectural components — task manager, programmer agent, and sandboxed execution — for deploying internal coding agents at scale.
+    blurb: Open-source SWE-agent framework built on LangGraph; covers core 
+      architectural components — task manager, programmer agent, and sandboxed 
+      execution — for deploying internal coding agents at scale.
     cluster: null
     tags:
       - coding-agent
@@ -2615,7 +2647,10 @@ resources:
     author: LangChain
     type: article
     license: null
-    blurb: Embedded interpreter runtimes let agents write code to coordinate tool calls, manage working state between steps, and control what gets surfaced into model context — reducing token pressure and enabling finer-grained orchestration than pure tool-dispatch.
+    blurb: Embedded interpreter runtimes let agents write code to coordinate 
+      tool calls, manage working state between steps, and control what gets 
+      surfaced into model context — reducing token pressure and enabling 
+      finer-grained orchestration than pure tool-dispatch.
     cluster: null
     tags:
       - sandboxing
@@ -2629,3 +2664,491 @@ resources:
     paywall: false
     superseded_by: null
     notes: null
+  - id: eval-smell-product-design
+    section: eval-philosophy
+    url: https://hamel.dev/blog/posts/eval-smell/
+    title: "\"It's Hard to Eval\" Is a Product Smell"
+    author: Hamel Husain
+    type: article
+    license: null
+    blurb: 'Difficulty eval-ing is a design flaw: artifacts hard to verify by machine
+      are hard to verify by users too. Three worked examples (data agent, PE curriculum
+      builder, workers-comp report) show how designing for human verifiability — provenance,
+      progressive disclosure, diff-scoped review — also makes automated evals tractable.'
+    cluster: null
+    tags:
+      - evals
+      - product-design
+      - verification
+      - provenance
+      - human-in-the-loop
+      - data-agents
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: cybersecurity-evals-patterns
+    section: eval-frameworks
+    url: https://eugeneyan.com//writing/cybersecurity-evals/
+    title: Patterns for Building Cybersecurity Evals
+    author: Eugene Yan
+    type: article
+    license: null
+    blurb: 'Four-component pattern for adversarial evals: sandboxed target, difficulty-modulating
+      inputs, tools, and a grader. Concrete design decisions for evaluating agents
+      on cybersecurity tasks.'
+    cluster: null
+    tags:
+      - evals
+      - security
+      - sandboxing
+      - grading
+      - adversarial
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: toctou-computer-use-agents
+    section: security
+    url: 
+      https://embracethered.com/blog/posts/2026/toctou-agent-what-you-click-is-not-what-you-get/
+    title: 'Computer-Use and TOCTOU: What You Click Is Not What You Get!'
+    author: embrace-the-red (Johann Rehberger)
+    type: article
+    license: null
+    blurb: 'TOCTOU race condition attack against computer-use agents: content shown
+      to the user at approval time can be swapped before the agent acts on it, breaking
+      human-in-the-loop trust assumptions. Reproduces the ChatGPT Operator vulnerability
+      disclosed by Jun Kokatsu with a full attack chain demo.'
+    cluster: null
+    tags:
+      - security
+      - computer-use
+      - toctou
+      - race-condition
+      - prompt-injection
+      - tool-misuse
+      - human-in-the-loop
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: cloudflare-workflows-saga-rollbacks
+    section: durable-execution
+    url: https://blog.cloudflare.com/rollbacks-for-workflows/
+    title: How we built saga rollbacks for Cloudflare Workflows
+    author: Cloudflare
+    type: article
+    license: null
+    blurb: 'Saga-pattern rollbacks in Cloudflare Workflows: each `step.do()` accepts
+      a compensating action, executed in reverse order on failure. Covers implementation
+      trade-offs, durability guarantees, and how compensation state is checkpointed
+      alongside forward progress.'
+    cluster: null
+    tags:
+      - durable-execution
+      - cloudflare-workflows
+      - saga-pattern
+      - rollbacks
+      - fault-tolerance
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: cloudflare-vulnerability-harness
+    section: coding-agent-infra
+    url: https://blog.cloudflare.com/build-your-own-vulnerability-harness/
+    title: Build your own vulnerability harness
+    author: Cloudflare
+    type: article
+    license: null
+    blurb: 'Multi-stage vulnerability discovery harness: state management, adversarial
+      review to suppress false positives, and strategies for routing around LLM context
+      limits in an automated triage loop.'
+    cluster: null
+    tags:
+      - harness-design
+      - security
+      - triage
+      - state-management
+      - context-limits
+      - false-positives
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: rlms-deep-agents-langchain
+    section: multi-agent-frameworks
+    url: https://www.langchain.com/blog/how-to-use-rlms-in-deep-agents
+    title: How to Use RLMs in Deep Agents
+    author: LangChain
+    type: article
+    license: null
+    blurb: Recursive language models (RLMs) address context rot by having agents
+      write code that fans out subagents over context chunks — grep/map/reduce 
+      patterns over large inputs. Benchmarked on OOLONG long-context reasoning; 
+      shows where turn-by-turn agents break down and dynamic subagent dispatch 
+      holds up.
+    cluster: null
+    tags:
+      - rlm
+      - subagents
+      - context-window
+      - fan-out
+      - long-context
+      - code-interpreter
+      - benchmarking
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: deep-agents-untrusted-code-wasm
+    section: sandboxing
+    url: 
+      https://www.langchain.com/blog/running-untrusted-agent-code-without-a-sandbox
+    title: How Deep Agents Run Untrusted Code Without a Sandbox
+    author: LangChain
+    type: article
+    license: null
+    blurb: In-process isolation via WASM + QuickJS as a lighter alternative to 
+      full VM/container sandboxes — covers least-privilege capability gating and
+      snapshot-based durable pauses for long-running agent tasks.
+    cluster: null
+    tags:
+      - sandboxing
+      - wasm
+      - code-execution
+      - isolation
+      - durable-execution
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: langchain-react-agent-benchmarking
+    section: eval-frameworks
+    url: https://www.langchain.com/blog/react-agent-benchmarking
+    title: Benchmarking Single Agent Performance
+    author: LangChain
+    type: article
+    license: null
+    blurb: Empirical study of how scaling instruction count and tool count 
+      degrades ReAct agent performance across claude-3.5-sonnet, gpt-4o, o1, and
+      o3-mini — two task domains, reproducible methodology.
+    cluster: null
+    tags:
+      - benchmarking
+      - react
+      - single-agent
+      - tool-use
+      - model-comparison
+      - langchain
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: langchain-openevals-agenteval
+    section: eval-frameworks
+    url: https://www.langchain.com/blog/evaluating-llms-with-openevals
+    title: Evaluating Large Language Models With OpenEvals
+    author: LangChain
+    type: article
+    license: null
+    blurb: Pre-built evaluators for LLM-as-judge, structured output correctness,
+      and agent trajectory assessment via OpenEvals and AgentEvals; covers 
+      scorer design, integration patterns, and production deployment.
+    cluster: null
+    tags:
+      - evals
+      - llm-as-judge
+      - agent-trajectory
+      - openevals
+      - langchain
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: designing-efficient-verifiers-legal-agents
+    section: eval-philosophy
+    url: 
+      https://www.langchain.com/blog/designing-efficient-verifiers-for-legal-agents
+    title: Designing Efficient Verifiers for Legal Agents
+    author: Harvey & LangChain Labs
+    type: article
+    license: null
+    blurb: Harvey + LangChain Labs study on building cheaper, more reliable 
+      LLM-as-judge verifiers for legal-agent eval and post-training signal — 
+      covering verifier calibration, cost-accuracy tradeoffs, and failure modes 
+      in domain-specific agentic pipelines.
+    cluster: null
+    tags:
+      - verifiers
+      - llm-as-judge
+      - eval
+      - legal
+      - post-training
+      - domain-specific
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: context-management-deep-agents
+    section: memory-systems
+    url: https://www.langchain.com/blog/context-management-for-deepagents
+    title: Context Management for Deep Agents
+    author: LangChain
+    type: article
+    license: null
+    blurb: 'How Deep Agents SDK handles context rot in long-running tasks: offloading,
+      summarization, and filesystem abstraction as first-class context management
+      primitives.'
+    cluster: null
+    tags:
+      - context-management
+      - memory
+      - summarization
+      - long-running-agents
+      - langchain
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: langchain-evals-deep-agents
+    section: eval-philosophy
+    url: https://www.langchain.com/blog/how-we-build-evals-for-deep-agents
+    title: How we build evals for Deep Agents
+    author: LangChain
+    type: article
+    license: null
+    blurb: Data curation, metrics selection, and testing strategies for 
+      evaluating deep/long-horizon agents — targeted behavioral measurement over
+      surface-level accuracy.
+    cluster: null
+    tags:
+      - evals
+      - deep-agents
+      - data-curation
+      - metrics
+      - langchain
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: langchain-fireworks-trace-judge
+    section: eval-philosophy
+    url: 
+      https://www.langchain.com/blog/building-a-100x-cheaper-trace-judge-with-fireworks
+    title: Building a 100x Cheaper Trace Judge with Fireworks
+    author: LangChain
+    type: article
+    license: null
+    blurb: Fine-tune a small open model on production trace signals to match 
+      frontier judge performance at ~1/100th the cost — covers data mining from 
+      traces, distillation pipeline, and eval results.
+    cluster: null
+    tags:
+      - llm-as-judge
+      - fine-tuning
+      - trace-mining
+      - cost-optimization
+      - production-evals
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: iterating-towards-llm-reliability-edd
+    section: eval-philosophy
+    url: 
+      https://www.langchain.com/blog/iterating-towards-llm-reliability-with-evaluation-driven-development
+    title: Iterating Towards LLM Reliability with Evaluation Driven Development
+    author: LangChain / Dosu
+    type: article
+    license: null
+    blurb: 'Evaluation-driven development (EDD) as a production discipline: build
+      eval suites before shipping, use production traces to seed regression datasets,
+      and gate iterative improvements on measurable score deltas. Practical case study
+      from Dosu at scale.'
+    cluster: null
+    tags:
+      - evaluation
+      - evaluation-driven-development
+      - langsmith
+      - production
+      - reliability
+      - tracing
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: aligning-llm-as-a-judge-human-preferences
+    section: eval-philosophy
+    url: 
+      https://www.langchain.com/blog/aligning-llm-as-a-judge-with-human-preferences
+    title: Aligning LLM-as-a-Judge with Human Preferences
+    author: LangChain
+    type: article
+    license: null
+    blurb: Few-shot calibration techniques for closing the gap between LLM-judge
+      scores and human preference labels — with concrete implementation in 
+      LangSmith's self-improving evaluators.
+    cluster: null
+    tags:
+      - llm-as-judge
+      - eval-calibration
+      - few-shot
+      - human-feedback
+      - langsmith
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: building-langgraph-agent-runtime
+    section: multi-agent-frameworks
+    url: https://www.langchain.com/blog/building-langgraph
+    title: 'Building LangGraph: Designing an Agent Runtime from First Principles'
+    author: LangChain
+    type: article
+    license: null
+    blurb: "First-principles design rationale behind LangGraph: why graph-based state
+      machines, how control flow and durability are handled, and what tradeoffs shaped
+      the runtime's architecture for production agents."
+    cluster: null
+    tags:
+      - langgraph
+      - agent-runtime
+      - graph-orchestration
+      - durability
+      - state-machines
+      - control-flow
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: claude-code-domain-specific-agent
+    section: coding-agent-infra
+    url: 
+      https://www.langchain.com/blog/how-to-turn-claude-code-into-a-domain-specific-coding-agent
+    title: How to turn Claude Code into a domain specific coding agent
+    author: LangChain
+    type: article
+    license: null
+    blurb: Context engineering techniques for specializing Claude Code toward 
+      domain-specific libraries — CLAUDE.md, tool allowlists, prompt injection —
+      with eval results comparing approaches.
+    cluster: null
+    tags:
+      - claude-code
+      - context-engineering
+      - domain-adaptation
+      - evals
+      - harness-design
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: langchain-evaluating-deep-agents
+    section: eval-philosophy
+    url: https://www.langchain.com/blog/evaluating-deep-agents-our-learnings
+    title: 'Evaluating Deep Agents: Our Learnings'
+    author: LangChain
+    type: article
+    license: null
+    blurb: 'Five evaluation patterns for long-horizon agents: bespoke unit tests,
+      single-step tool validation, full-turn checks, multi-turn simulations, and environment
+      setup — grounded in production experience.'
+    cluster: null
+    tags:
+      - evals
+      - deep-agents
+      - multi-turn
+      - simulation
+      - langchain
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: langchain-choose-right-sandbox-agents
+    section: sandboxing
+    url: 
+      https://www.langchain.com/blog/how-to-choose-the-right-sandbox-for-your-agent
+    title: How to Choose the Right Sandbox for AI Agents
+    author: LangChain
+    type: article
+    license: null
+    blurb: 'Decision framework for sandbox selection: filesystem isolation, network
+      access controls, resource limits, and microVM trade-offs for code-executing
+      agents.'
+    cluster: null
+    tags:
+      - sandboxing
+      - microvm
+      - isolation
+      - code-execution
+      - security
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
+  - id: langgraph-fault-tolerance
+    section: durable-execution
+    url: https://www.langchain.com/blog/fault-tolerance-in-langgraph
+    title: 'Fault Tolerance in LangGraph: Retries, Timeouts and Error Handlers'
+    author: LangChain
+    type: article
+    license: null
+    blurb: Three fault-tolerance primitives inside LangGraph — `RetryPolicy` 
+      (backoff retries), `TimeoutPolicy` (wall-clock and idle caps), and 
+      `error_handler` (post-retry cleanup) — and how they compose. Applies the 
+      SAGA pattern to multi-step workflows with real-world side effects.
+    cluster: null
+    tags:
+      - langgraph
+      - fault-tolerance
+      - retry
+      - timeout
+      - saga
+      - error-handling
+      - workflows
+    added_at: '2026-07-02'
+    verified_at: null
+    archived: false
+    paywall: false
+    superseded_by: null
+    notes: null
diff --git a/scout/seen.yaml b/scout/seen.yaml
index 1916c7f..b8d8a71 100644
--- a/scout/seen.yaml
+++ b/scout/seen.yaml
@@ -153,3 +153,461 @@ seen:
     title: Mission Control for Self-Hosted LangSmith on Kubernetes
     source_id: langchain-blog
     rejected_at: '2026-05-28'
+  - url: https://blog.cloudflare.com/monetization-gateway/
+    title: 'Announcing the Monetization Gateway: charge for any resource behind Cloudflare
+      via x402'
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: https://blog.cloudflare.com/agentic-internet-bot-report/
+    title: 'Content Independence Day, one year on: building the business model for
+      the agentic Internet'
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: https://blog.cloudflare.com/making-ai-search-smarter/
+    title: Making AI search smarter
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: https://blog.cloudflare.com/content-independence-day-ai-options/
+    title: 'Your site, your rules: new AI traffic options for all customers'
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: https://blog.cloudflare.com/attribution-business-insights/
+    title: Unmasking the crawls with Attribution Business Insights
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: https://blog.cloudflare.com/oauth-for-all/
+    title: Unlocking the Cloudflare app ecosystem with OAuth for all
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: https://blog.cloudflare.com/post-quantum-eo-2026/
+    title: The White House's post-quantum executive order is an important 
+      milestone. It’s time to get to work
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: https://blog.cloudflare.com/hyper-bug/
+    title: How we found a bug in the hyper HTTP library
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: https://blog.cloudflare.com/temporary-accounts/
+    title: Temporary Cloudflare Accounts for AI agents
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: https://blog.cloudflare.com/celebrating-12-years-of-project-galileo/
+    title: Celebrating 12 years of Project Galileo
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: https://blog.cloudflare.com/agents-platform-flue-sdk/
+    title: Bringing more agent harnesses and frameworks to Cloudflare, starting 
+      with Flue
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: https://blog.cloudflare.com/cloudflare-one-stack/
+    title: 'Introducing the Cloudflare One stack: agent-powered deployment'
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: https://blog.cloudflare.com/dmarc-management-ga/
+    title: Cloudflare DMARC Management is now generally available
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: https://blog.cloudflare.com/ensemble-ai-talent-joins-cloudflare/
+    title: Growing the Cloudflare AI team with talent from Ensemble AI
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: https://blog.cloudflare.com/scaling-security-scans/
+    title: 'Scaling Security Insights: how we achieved a 10x increase in global scanning
+      capacity'
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: https://blog.cloudflare.com/private-origins-dns-routing/
+    title: Route public traffic to private applications with Cloudflare
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: https://blog.cloudflare.com/frontier-model-defense/
+    title: "Defend against frontier cyber models: Cloudflare's architecture as customer
+      zero"
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: https://blog.cloudflare.com/realtime-threat-intel-waf-rules/
+    title: Turning Cloudflare’s threat indicators into real-time WAF rules
+    source_id: cloudflare-ai-agents
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/introducing-openwiki-an-open-source-agent-for-repo-documentation
+    title: 'OpenWiki: Open Source Repo Documentation for Coding Agents'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/how-pendo-used-langsmith-to-trace-novus-from-user-behavior-to-code-fixes
+    title: How Pendo uses LangSmith to trace Novus from user behavior to code 
+      fixes
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/wiki-memory
+    title: ''
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/unified-stack-for-evaluating-agents
+    title: 'Harbor x LangChain: A Unified Stack for Evaluating Agents'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/how-rippling-went-ai-native-across-every-product-in-6-months-with-deep-agents-and-langsmith
+    title: How Rippling built production AI in 6 months with Deep Agents and 
+      LangSmith
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/production-monitoring
+    title: 'Agent Observability: How to Monitor and Evaluate LLM Agents in Production'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/extraction-benchmarking
+    title: Extraction Benchmarking
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/benchmarking-agent-tool-use
+    title: Benchmarking Agent Tool Use
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/high-cardinality
+    title: Benchmarking Query Analysis in High Cardinality Situations
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/pairwise-evaluations-with-langsmith
+    title: Pairwise Evaluations with LangSmith
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/agent-observability-powers-agent-evaluation
+    title: How to Debug & Evaluate AI Agents with Observability — LangChain 
+      Guide
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/evaluating-skills
+    title: Evaluating Skills
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/agent-evaluation-readiness-checklist
+    title: Agent Evaluation Readiness Checklist
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/introducing-dynamic-subagents-in-deep-agents
+    title: Introducing Dynamic Subagents in Deep Agents
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/how-candidly-built-state-aware-agent-harnesses-with-langsmith
+    title: ''
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/deep-agents-prompt-caching
+    title: Prompt Caching with Deep Agents
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/june-2026-langchain-newsletter
+    title: 'June 2026: LangChain Newsletter — Fleet On-Call Copilot, Deep Agents Rubrics,
+      and More'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/full-text-search-in-smithdb-constructing-and-querying-our-inverted-index-pt-2
+    title: How we built SmithDB’s inverted index for full-text search
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/why-the-best-agents-are-simpler-than-you-think-sierra-max-agency-podcast
+    title: 'Why the Best AI Agents Are Simple: Sierra’s Zack Reneau-Wedeen on the
+      Max Agency Podcast'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/the-art-of-loop-engineering
+    title: The Art of Loop Engineering
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/give-your-ai-agent-its-own-computer
+    title: Give your agent its own computer
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/introducing-rubrics-for-deepagents
+    title: 'Introducing Rubrics: Build Agents that Evaluate and Correct Their Work'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/langsmith-langchain-oss-eu-ai-act
+    title: How LangSmith and LangChain OSS Help You Meet EU AI Act Requirements
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/agentic-engineering-redefining-software-engineering
+    title: 'Agentic Engineering: How Swarms of AI Agents Are Redefining Software Engineering'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/customers-klarna
+    title: How Klarna's AI assistant redefined customer support at scale for 85 
+      million active users
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/how-to-give-your-agent-memory
+    title: How to Build Memory into AI Agents
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/langsmith-agent-builder
+    title: Introducing LangSmith’s No Code Agent Builder
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/why-fleet-has-both-general-purpose-chat-and-specialized-agents
+    title: Why Fleet Has General Purpose Chat and Specialized Agents
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/reusable-langsmith-evaluator-templates
+    title: Reusable Evaluators and Evaluator Templates in LangSmith
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/human-judgment-in-the-agent-improvement-loop
+    title: Human judgment in the agent improvement loop
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/testing-fine-tuned-open-source-models-in-langsmith
+    title: Testing Fine Tuned Open Source Models in LangSmith
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/public-langsmith-benchmarks
+    title: Sharing LangSmith Benchmarks
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/customers-factory
+    title: How Factory used LangSmith to automate their feedback loop and 
+      improve iteration speed by 2x
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/langsmith-homepage-redesign-and-resource-tags
+    title: 'LangSmith: Redesigned product homepage and Resource Tags for better organization'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/introducing-open-swe-an-open-source-asynchronous-coding-agent
+    title: 'Introducing Open SWE: An Open-Source Asynchronous Coding Agent'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/customers-monte-carlo
+    title: 'Monte Carlo: Building Data + AI Observability Agents with LangGraph and
+      LangSmith'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/agent-engineering-a-new-discipline
+    title: 'Agent Engineering: A New Discipline'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/how-we-made-coding-agent-spend-predictable
+    title: How LangChain Made Coding Agent Spend Predictable
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/pinecone-serverless
+    title: Build and deploy a RAG app with Pinecone Serverless
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/test-run-comparisons
+    title: Test Run Comparisons
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/announcing-langsmith-is-now-a-transactable-offering-in-the-azure-marketplace
+    title: Announcing LangSmith is now a transactable offering in the Azure 
+      Marketplace
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/regression-testing
+    title: Regression Testing with LangSmith
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/what-is-an-agent
+    title: What is an AI agent?
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/customers-new-computer
+    title: 'Improving Memory Retrieval: How New Computer achieved 50% higher recall
+      with LangSmith'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/customers-replit
+    title: Pushing LangSmith to new limits with Replit Agent's complex workflows
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/promptim
+    title: 'Promptim: an experimental library for prompt optimization'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/opentelemetry-langsmith
+    title: Introducing OpenTelemetry support for LangSmith
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/end-to-end-opentelemetry-langsmith
+    title: Introducing End-to-End OpenTelemetry Support in LangSmith
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/how-to-think-about-agent-frameworks
+    title: How to think about agent frameworks
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/interrupt-2025-recap
+    title: 'Recap of Interrupt 2025: The AI Agent Conference by LangChain'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/how-and-when-to-build-multi-agent-systems
+    title: How and when to build multi-agent systems
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/introducing-align-evals
+    title: 'Introducing Align Evals: Streamlining LLM Application Evaluation'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/on-agent-frameworks-and-agent-observability
+    title: On Agent Frameworks and Agent Observability
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/customers-monday
+    title: 'monday Service + LangSmith: Building a Code-First Evaluation Strategy
+      from Day 1'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/how-we-built-langchains-gtm-agent
+    title: How we built LangChain’s GTM Agent
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/building-box-ai-how-an-enterprise-content-platform-went-ai-native-with-deep-agents
+    title: How Box AI built enterprise content agents with Deep Agents
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/benchling-max-agency-podcast
+    title: How Benchling builds agents when the smartest AI isn't smart enough
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/agents-and-applications
+    title: The Missing Link Between Agents and Applications
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/full-text-search-in-smithdb-designing-an-inverted-index-for-object-storage
+    title: 'Full Text Search in SmithDB: Designing an Inverted Index for Object Storage'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/model-neutrality
+    title: 'Model Neutrality: Why Avoiding AI Vendor Lock-In Matters'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/how-to-build-a-custom-agent-harness
+    title: How to Build a Custom Agent Harness
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/how-harmonic-rebuilt-scout-on-deep-agents-and-4xd-retention-with-langsmith
+    title: ''
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://www.langchain.com/blog/interpreter-skills
+    title: 'Interpreter Skills: Building Workflows for Agents'
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: 
+      https://www.langchain.com/blog/financial-ai-that-investigates-macro-trends-eu-economic-analysis-with-you-com-and-langchain
+    title: ''
+    source_id: langchain-blog
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/how-chatgpt-adoption-has-expanded
+    title: How ChatGPT adoption has expanded
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/genebench-pro/case-studies
+    title: Inside Genebench-Pro
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/introducing-genebench-pro
+    title: Introducing GeneBench-Pro
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/core-dump-epidemiology-data-infrastructure-bug
+    title: 'Core dump epidemiology: fixing an 18-year-old bug'
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/mapping-ai-jobs-transition-eu
+    title: Mapping Europe’s AI Workforce Opportunity
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/hp-frontier-partnership
+    title: HP Inc. launches Frontier strategic partnership with OpenAI
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/previewing-gpt-5-6-sol
+    title: 'Previewing GPT-5.6 Sol: a next-generation model'
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/how-agents-are-transforming-work
+    title: How agents are transforming work
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/openai-broadcom-jalapeno-inference-chip
+    title: OpenAI and Broadcom unveil LLM-optimized inference chip
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/helping-build-shared-standards-for-advanced-ai
+    title: Helping build shared standards for advanced AI
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/gpt-5-immunology-mystery
+    title: How GPT-5 helped immunologist Derya Unutmaz solve a 3-year-old 
+      mystery
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/omio
+    title: How Omio is building the future of conversational travel
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/patch-the-planet
+    title: 'Patch the Planet: a Daybreak initiative to support open source maintainers'
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/daybreak-securing-the-world
+    title: 'Daybreak: Tools for securing every organization in the world'
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/codex-maxxing-long-running-work
+    title: Codex-maxxing for long-running work
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/samsung-electronics-chatgpt-codex-deployment
+    title: Samsung Electronics brings ChatGPT and Codex to employees
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/chatgpt-enterprise-spend-controls
+    title: New usage analytics and updated spend controls for enterprises
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/improving-health-intelligence-in-chatgpt
+    title: Improving health intelligence in ChatGPT
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/diagnose-rare-childhood-diseases
+    title: Using AI to help physicians diagnose rare genetic diseases affecting 
+      children
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/ai-chemist-improves-reaction
+    title: A near-autonomous AI chemist improves a challenging reaction in 
+      medicinal chemistry
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://openai.com/index/introducing-life-sci-bench
+    title: Introducing LifeSciBench
+    source_id: openai-news
+    rejected_at: '2026-07-02'
+  - url: https://zed.dev/blog/hidden-gems-part-4
+    title: 'Hidden Gems: Part 4'
+    source_id: zed-blog
+    rejected_at: '2026-07-02'
diff --git a/sources.yaml b/sources.yaml
index 856b325..d373301 100644
--- a/sources.yaml
+++ b/sources.yaml
@@ -5,7 +5,7 @@ sources:
     type: atom
     url: https://simonwillison.net/tags/ai-agents/atom
     cadence: weekly
-    last_checked_at: 2026-05-28
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -13,7 +13,7 @@ sources:
     type: rss
     url: https://hamel.dev/index.xml
     cadence: weekly
-    last_checked_at: 2026-05-28
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -21,7 +21,7 @@ sources:
     type: rss
     url: https://eugeneyan.com/rss/
     cadence: weekly
-    last_checked_at: 2026-05-28
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -29,7 +29,7 @@ sources:
     type: rss
     url: https://huyenchip.com/feed.xml
     cadence: weekly
-    last_checked_at: 2026-05-28
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -37,7 +37,7 @@ sources:
     type: rss
     url: https://embracethered.com/blog/index.xml
     cadence: weekly
-    last_checked_at: 2026-05-28
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -45,7 +45,7 @@ sources:
     type: rss
     url: https://blog.cloudflare.com/tag/ai-agents/rss
     cadence: weekly
-    last_checked_at: 2026-05-28
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -53,7 +53,7 @@ sources:
     type: rss
     url: https://www.langchain.com/blog/rss.xml
     cadence: weekly
-    last_checked_at: 2026-05-28
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -61,7 +61,7 @@ sources:
     type: rss
     url: https://cognition.ai/rss.xml
     cadence: weekly
-    last_checked_at: 2026-05-28
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -71,7 +71,7 @@ sources:
     type: github-releases
     url: https://github.com/langchain-ai/langgraph/releases.atom
     cadence: weekly
-    last_checked_at: 2026-05-28
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -79,7 +79,7 @@ sources:
     type: github-releases
     url: https://github.com/openai/openai-agents-python/releases.atom
     cadence: weekly
-    last_checked_at: 2026-05-28
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -87,7 +87,7 @@ sources:
     type: github-releases
     url: https://github.com/UKGovernmentBEIS/inspect_ai/releases.atom
     cadence: weekly
-    last_checked_at: 2026-05-28
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -95,7 +95,7 @@ sources:
     type: github-releases
     url: https://github.com/letta-ai/letta/releases.atom
     cadence: weekly
-    last_checked_at: 2026-05-28
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -103,7 +103,7 @@ sources:
     type: github-releases
     url: https://github.com/microsoft/autogen/releases.atom
     cadence: weekly
-    last_checked_at: 2026-05-28
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -111,7 +111,7 @@ sources:
     type: github-releases
     url: https://github.com/google/adk-python/releases.atom
     cadence: weekly
-    last_checked_at: 2026-05-28
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -119,7 +119,7 @@ sources:
     type: github-releases
     url: https://github.com/huggingface/smolagents/releases.atom
     cadence: weekly
-    last_checked_at: 2026-05-28
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -127,7 +127,7 @@ sources:
     type: github-releases
     url: https://github.com/modelcontextprotocol/servers/releases.atom
     cadence: weekly
-    last_checked_at: 2026-05-28
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -137,7 +137,7 @@ sources:
     type: atom
     url: https://vercel.com/atom
     cadence: weekly
-    last_checked_at: 2026-06-16
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -145,7 +145,7 @@ sources:
     type: rss
     url: https://openai.com/news/rss.xml
     cadence: weekly
-    last_checked_at: 2026-06-16
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -153,7 +153,7 @@ sources:
     type: rss
     url: https://zed.dev/blog.rss
     cadence: weekly
-    last_checked_at: 2026-06-16
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -161,7 +161,7 @@ sources:
     type: github-releases
     url: https://github.com/browser-use/browser-use/releases.atom
     cadence: weekly
-    last_checked_at: 2026-06-16
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -169,7 +169,7 @@ sources:
     type: github-releases
     url: https://github.com/microsoft/playwright/releases.atom
     cadence: weekly
-    last_checked_at: 2026-06-16
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null
 
@@ -177,6 +177,6 @@ sources:
     type: github-releases
     url: https://github.com/browserbase/stagehand/releases.atom
     cadence: weekly
-    last_checked_at: 2026-06-16
+    last_checked_at: 2026-07-02
     enabled: true
     notes: null