{ "version": "0.1.0", "generatedAt": "2026-05-16T02:08:45.557Z", "purpose": "Parsed article index generated from file-backed Markdown sources and visual sidecars for article list and Command-K search surfaces.", "sourceDir": "content/articles", "articleCount": 7, "sourceFileCount": 7, "visualSidecarCount": 7, "articles": [ { "slug": "building-a-useful-ai-leaderboard-without-fooling-ourselves", "title": "Building a Useful AI Leaderboard Without Fooling Ourselves", "category": "Benchmarks", "description": "A field guide for ranking AI systems without being fooled by contamination, brittle prompts, saturated tests, or one-number theatre.", "date": "15 May 2026", "readingTime": "12 min read", "reviewer": "Saujas", "status": "Published draft", "canonicalPath": "/articles/building-a-useful-ai-leaderboard-without-fooling-ourselves", "sourcePath": "content/articles/building-a-useful-ai-leaderboard-without-fooling-ourselves.md", "visualPath": "content/articles/building-a-useful-ai-leaderboard-without-fooling-ourselves.visual.json", "publicDraftPath": "/reports/editorial/drafts/building-a-useful-ai-leaderboard-without-fooling-ourselves.md", "publicVisualPacketPath": "/reports/editorial/visual-packets/building-a-useful-ai-leaderboard-without-fooling-ourselves.json", "sourceCount": 5, "visualModuleCount": 6 }, { "slug": "mechanistic-interpretability-for-operators-not-mystics", "title": "Mechanistic Interpretability for Operators, Not Mystics", "category": "Mechanistic Interpretability", "description": "A practical explanation of circuits, sparse autoencoders, feature dashboards, and why interpretability should become part of AI system audits.", "date": "16 May 2026", "readingTime": "14 min read", "reviewer": "Sanjay Prasad", "status": "Published draft", "canonicalPath": "/articles/mechanistic-interpretability-for-operators-not-mystics", "sourcePath": "content/articles/mechanistic-interpretability-for-operators-not-mystics.md", "visualPath": "content/articles/mechanistic-interpretability-for-operators-not-mystics.visual.json", "publicDraftPath": "/reports/editorial/drafts/mechanistic-interpretability-for-operators-not-mystics.md", "publicVisualPacketPath": "/reports/editorial/visual-packets/mechanistic-interpretability-for-operators-not-mystics.json", "sourceCount": 4, "visualModuleCount": 6 }, { "slug": "cost-curves-for-frontier-reasoning-models", "title": "Cost Curves for Frontier Reasoning Models", "category": "Model Economics", "description": "A buyer's framework for converting token prices into workflow cost, including reasoning tokens, cache hits, batch discounts, and tool calls.", "date": "17 May 2026", "readingTime": "13 min read", "reviewer": "Saujas", "status": "Published draft", "canonicalPath": "/articles/cost-curves-for-frontier-reasoning-models", "sourcePath": "content/articles/cost-curves-for-frontier-reasoning-models.md", "visualPath": "content/articles/cost-curves-for-frontier-reasoning-models.visual.json", "publicDraftPath": "/reports/editorial/drafts/cost-curves-for-frontier-reasoning-models.md", "publicVisualPacketPath": "/reports/editorial/visual-packets/cost-curves-for-frontier-reasoning-models.json", "sourceCount": 4, "visualModuleCount": 6 }, { "slug": "agent-benchmarks-that-survive-real-work", "title": "Agent Benchmarks That Survive Real Work", "category": "Agents", "description": "A map of coding, terminal, browser, OS, and customer-support agent benchmarks, and what each misses when used alone.", "date": "18 May 2026", "readingTime": "15 min read", "reviewer": "Saujas", "status": "Published draft", "canonicalPath": "/articles/agent-benchmarks-that-survive-real-work", "sourcePath": "content/articles/agent-benchmarks-that-survive-real-work.md", "visualPath": "content/articles/agent-benchmarks-that-survive-real-work.visual.json", "publicDraftPath": "/reports/editorial/drafts/agent-benchmarks-that-survive-real-work.md", "publicVisualPacketPath": "/reports/editorial/visual-packets/agent-benchmarks-that-survive-real-work.json", "sourceCount": 4, "visualModuleCount": 3 }, { "slug": "prompt-caching-batch-api-and-the-real-ai-cost-stack", "title": "Prompt Caching, Batch API, and the Real AI Cost Stack", "category": "Model Economics", "description": "How prompt caching, batch execution, retries, tool calls, and review loops change the economics of production AI workloads.", "date": "19 May 2026", "readingTime": "11 min read", "reviewer": "Saujas", "status": "Published draft", "canonicalPath": "/articles/prompt-caching-batch-api-and-the-real-ai-cost-stack", "sourcePath": "content/articles/prompt-caching-batch-api-and-the-real-ai-cost-stack.md", "visualPath": "content/articles/prompt-caching-batch-api-and-the-real-ai-cost-stack.visual.json", "publicDraftPath": "/reports/editorial/drafts/prompt-caching-batch-api-and-the-real-ai-cost-stack.md", "publicVisualPacketPath": "/reports/editorial/visual-packets/prompt-caching-batch-api-and-the-real-ai-cost-stack.json", "sourceCount": 4, "visualModuleCount": 3 }, { "slug": "designing-the-indian-enterprise-ai-workflow-benchmark", "title": "Designing the Indian Enterprise AI Workflow Benchmark", "category": "Indian Workflows", "description": "A concrete v0.1 benchmark plan for Indian business workflows: data shape, scoring, task refresh, and buyer-facing outputs.", "date": "20 May 2026", "readingTime": "16 min read", "reviewer": "Saujas", "status": "Published draft", "canonicalPath": "/articles/designing-the-indian-enterprise-ai-workflow-benchmark", "sourcePath": "content/articles/designing-the-indian-enterprise-ai-workflow-benchmark.md", "visualPath": "content/articles/designing-the-indian-enterprise-ai-workflow-benchmark.visual.json", "publicDraftPath": "/reports/editorial/drafts/designing-the-indian-enterprise-ai-workflow-benchmark.md", "publicVisualPacketPath": "/reports/editorial/visual-packets/designing-the-indian-enterprise-ai-workflow-benchmark.json", "sourceCount": 4, "visualModuleCount": 3 }, { "slug": "open-weight-inference-economics-for-enterprise-ai", "title": "Open-Weight Inference Economics for Enterprise AI", "category": "Model Economics", "description": "A buyer's map for deciding when Mistral, DeepSeek, Qwen, and other open-weight routes belong beside managed frontier APIs.", "date": "21 May 2026", "readingTime": "17 min read", "reviewer": "Saujas", "status": "Published draft", "canonicalPath": "/articles/open-weight-inference-economics-for-enterprise-ai", "sourcePath": "content/articles/open-weight-inference-economics-for-enterprise-ai.md", "visualPath": "content/articles/open-weight-inference-economics-for-enterprise-ai.visual.json", "publicDraftPath": "/reports/editorial/drafts/open-weight-inference-economics-for-enterprise-ai.md", "publicVisualPacketPath": "/reports/editorial/visual-packets/open-weight-inference-economics-for-enterprise-ai.json", "sourceCount": 5, "visualModuleCount": 6 } ] }