Context files
The shelf: every file an agent can load, filed under the concept it belongs to and sized in tokens.
Context files is for loading: each row below is a plain markdown file — a research report, or one concept's facts — and the link on it is the file itself. Click a root or a concept name to expand it in place; nothing here navigates except the file links.Concepts is the other tab: the same research as a map you browse.
A context file is a document an agent loads whole: a research report on one subject, written to be read into a context window rather than browsed. Third-party full text is never republished here — see the ethos. Afacts file is one concept's facets and facts as a markdown list, each fact carrying the URL it came from. How these files are filed and formatted. Agents: fetch /context.md for this whole listing as one file; any row below is the second and last hop.
323 files · 4.9 MB · 467 concepts · 38,738 facts in 6,889 facets · 10 roots · generated 2026-10-01
AI & LLM Engineering75 files · 115 concepts
- A2A Protocol Interoperability2 KB · ~578 tokens.mdReference for the Agent-to-Agent protocol and cross-framework agent communication.cited by A2A Protocol Interoperability
- Agent Council1 KB · ~228 tokens.mdCollect multiple AI opinions and synthesize one answer.cited by Agent Council
- Agent Harness Construction2 KB · ~495 tokens.md当你在改进智能体的规划、调用工具、从错误中恢复以及收敛到完成状态的方式时，使用此技能。cited by Agent Harness Construction
- Agent Identity, Authorization & Payments24 KB · ~6,073 tokens.mdThe trust / permission / value layer for autonomous AI agents acting withcited by Agent Identity, Authorization & Payments
- Agent Plan Writing3 KB · ~644 tokens.mdAgent plan writing is the discipline of designing execution plans for AI agent workflows. The harness matters more than the model. Agent completion rates depend more on action-space design, context encited by Agent Plan Writing
- Agent Runtime Sandboxes & Code Execution23 KB · ~5,886 tokens.mdSecure, ephemeral cloud environments where an AI agent runs LLM-generated code, uses a computer,cited by Agent Runtime Sandboxes & Code Execution
- Agentic RL — Reinforcement Learning for LLM Agents28 KB · ~7,170 tokens.mdThe 2024–2026 frontier discipline of training an LLM to act over many turns — calling tools, searching, browsing, editing code, driving a computer — by optimizing the whole multi-step trajectory againcited by Agentic RL — Reinforcement Learning for LLM Agents
- AI Agent Ecosystems2 KB · ~468 tokens.mdComprehensive reference for AI agent development, orchestration, infrastructure, and security.cited by AI Agent Ecosystems
- AI Datastores2 KB · ~435 tokens.md| Need | Recommended |cited by AI Datastores
- AI Gateways & LLM Proxy Infrastructure20 KB · ~5,175 tokens.mdAn AI gateway (LLM gateway / LLM proxy) is the production traffic-and-controlcited by AI Gateways & LLM Proxy Infrastructure
- AI Programming Languages1 KB · ~365 tokens.md| Scenario | Language | Framework |cited by AI Programming Languages
- Aider in Open-Source Agentic Coding with Ollama24 KB · ~6,133 tokens.mdAider (aider.chat / github.com/Aider-AI/aider) is a mature, git-native CLI pair-programming tool, not a fully autonomous agent. It has no native tool-calling or MCP (Model Context Protocol, the standacited by Aider in Open-Source Agentic Coding with Ollama
- ASUS NUC BIOS Thunderbolt options and the iSetupCfg CLI20 KB · ~5,235 tokens.mdWhat is and is not known about the firmware and BIOS of an ASUS NUC 15 Pro for Thunderbolt eGPU use on Linux — the options that could govern the tunnel and the evidence for each, the undocumented-opticited by ASUS NUC BIOS Thunderbolt options and the iSetupCfg CLI
- Autonomous Loop Patterns24 KB · ~6,149 tokens.md> 兼容性说明 (v1.8.0): autonomous-loops 保留一个发布周期。cited by Autonomous Loop Patterns
- Blackwell sm_120 local-LLM inference stack on Linux: llama.cpp, Ollama, PyTorch and vLLM over a Thunderbolt eGPU45 KB · ~11,590 tokens.mdBuilding and choosing the local-LLM inference stack for a consumer Blackwell GPU (sm_120, RTX 5080 16 GB) on Linux — the llama.cpp CUDA build matrix, Ollama's bundled CUDA backends and GPU discovery,cited by Blackwell sm_120 local-LLM inference stack on Linux: llama.cpp, Ollama, PyTorch and vLLM over a Thunderbolt eGPU
- Claude Code Plugins3 KB · ~773 tokens.mdReference for the Claude Code extension system: plugins, hooks, commands, and agents. Backed by references/claude-code-plugins-context.md.cited by Claude Code Plugins
- Claude Code Skills3 KB · ~756 tokens.mdReference for the Claude Code skills ecosystem — anatomy, authoring, discovery, distribution, management, composition, and optimization. Backed by references/claude-code-skills-context.md.cited by Claude Code Skills
- Conceptual Family Exploration18 KB · ~4,645 tokens.mdGiven a subject, find the useful, relevant, novel, and interesting concepts incited by Conceptual Family Exploration
- Continuous Learning System11 KB · ~2,705 tokens.mdClaude Codeセッションを信頼度スコアリング付きの小さな学習済み行動である「インスティンクト」を通じて再利用可能な知識に変える高度な学習システム。cited by Continuous Learning System
- Cross-Encoder Re-ranking vs Bi-Encoders for Terminal Validation in RAG and Deduplication12 KB · ~3,092 tokens.mdThis report investigates the comparative advantages and architectural roles of Cross-Encoders and Bi-Encoders in modern RAG (Retrieval-Augmented Generation) systems, specifically focusing on terminalcited by Cross-Encoder Re-ranking vs Bi-Encoders for Terminal Validation in RAG and Deduplication
- Deep Research13 KB · ~3,387 tokens.mdYou are a research analyst. Produce cited research reports from multiple web sources using firecrawl and exa MCP tools. Deliver inline-cited findings organized by theme, with an executive summary, soucited by Deep Research
- Deep Research Methods14 KB · ~3,696 tokens.mdMethodology reference for rigorous AI-agent research. Complements deep-research (tool usage for firecrawl/exa) with research thinking — how to decompose questions, evaluate sources, synthesize findingcited by Deep Research Methods
- Diagnosing a GPU that has fallen off the bus: config space, MMIO chip ID and AER44 KB · ~11,207 tokens.mdWhy an NVIDIA GPU in a Thunderbolt or USB4 enclosure reports 'fallen off the bus' on Linux — eleven ranked root causes with the observable that separates each, a triage decision tree, a read-only captcited by Diagnosing a GPU that has fallen off the bus: config space, MMIO chip ID and AER
- Diffusion & Generative-Media Models11 KB · ~2,844 tokens.mdThe model family that generates continuous media (images, video, audio) by learning to reverse a noising process. A model-layer reference under the ai-agent-engineering hub (2024–2026). This is the gecited by Diffusion & Generative-Media Models
- Distributed Training & Training Infrastructure7 KB · ~1,833 tokens.md> Hub reference under ai-agent-engineering (hub-and-spoke). Owns the LLM training-infrastructure layer: how you split a model + optimizer + activations across many GPUs to TRAIN it. Loaded on demand wcited by Distributed Training & Training Infrastructure
- Durable Agent Execution & Long-Running Agent Runtimes22 KB · ~5,563 tokens.mdThe infrastructure/platform layer that lets AI agents run for minutes, hours,cited by Durable Agent Execution & Long-Running Agent Runtimes
- Extraction Evaluation Frameworks10 KB · ~2,458 tokens.mdThe transition from traditional Optical Character Recognition (OCR) to Large Language Model (LLM)-driven document extraction has fundamentally altered the landscape of automated data extraction. Whilecited by Extraction Evaluation Frameworks
- GPU & Accelerator Kernels for LLMs31 KB · ~7,960 tokens.mdPROVENANCEcited by GPU & Accelerator Kernels for LLMs
- GPU hot-unplug and surprise-removal safety for a Thunderbolt eGPU on Linux22 KB · ~5,644 tokens.mdWhy unplugging a Thunderbolt eGPU is unsafe on Linux with NVIDIA and what the kernel does about it — pciehp surprise link-down handling, the PCI core's disconnected flag and all-ones MMIO, DPC versuscited by GPU hot-unplug and surprise-removal safety for a Thunderbolt eGPU on Linux
- Health monitoring, alerting and safe automated recovery for a Thunderbolt eGPU43 KB · ~11,107 tokens.mdWatching a Thunderbolt eGPU on a headless Linux box and recovering it safely — the kernel-log and sysfs signatures worth alerting on, a guarded liveness probe that tells a dead GPU from a bridge withcited by Health monitoring, alerting and safe automated recovery for a Thunderbolt eGPU
- Heuristic Boilerplate Removal & Text-Density Algorithms for Web Extraction9 KB · ~2,402 tokens.mdHeuristic boilerplate removal and text-density algorithms are foundational techniques used in web scraping and natural language processing to separate primary article content from peripheral noise (bocited by Heuristic Boilerplate Removal & Text-Density Algorithms for Web Extraction
- Hybrid graphics on Linux: Intel iGPU plus NVIDIA eGPU for headless compute19 KB · ~4,833 tokens.mdRunning an Intel integrated GPU alongside an NVIDIA Thunderbolt eGPU used only for compute on Linux — xe versus i915 binding on Arrow Lake, how the desktop picks a primary GPU, keeping GNOME off the ecited by Hybrid graphics on Linux: Intel iGPU plus NVIDIA eGPU for headless compute
- Hybrid Score Fusion (Reciprocal Rank Fusion - RRF)6 KB · ~1,418 tokens.mdReciprocal Rank Fusion (RRF) is the industry standard zero-shot method for combining ranked lists in hybrid search pipelines (e.g., merging sparse BM25 and dense vector results). It computes a unifiedcited by Hybrid Score Fusion (Reciprocal Rank Fusion - RRF)
- Idle power and energy accounting for an always-on Thunderbolt eGPU27 KB · ~6,784 tokens.mdWhere an always-on Thunderbolt NVIDIA eGPU box spends its idle watts, how to measure wall, GPU and host power honestly, which levers (persistence mode, runtime D3, power limits, keep-alive policy) tracited by Idle power and energy accounting for an always-on Thunderbolt eGPU
- Iterative Retrieval7 KB · ~1,850 tokens.mdマルチエージェントワークフローにおける「コンテキスト問題」を解決します。サブエージェントは作業を開始するまで、どのコンテキストが必要かわかりません。cited by Iterative Retrieval
- Linux thunderbolt driver: host_reset, CLx and bolt authorization49 KB · ~12,525 tokens.mdHow a Thunderbolt 4 or USB4 host builds the PCIe tunnel to an eGPU enclosure on Linux and what governs it — host router, connection manager, retimers and the enclosure switch; the 32 Gb/s tunnel ceilicited by Linux thunderbolt driver: host_reset, CLx and bolt authorization
- LlamaParse vs. Docling for LLM Ingestion5 KB · ~1,408 tokens.mdThis report evaluates LlamaParse and Docling, two leading layout-aware document parsers designed for Large Language Model (LLM) ingestion and Retrieval-Augmented Generation (RAG) pipelines. Both toolscited by LlamaParse vs. Docling for LLM Ingestion
- LLM Fine-Tuning & PEFT31 KB · ~7,811 tokens.mdAdapting a pretrained LLM to a specific task, domain, format, or behavior bycited by LLM Fine-Tuning & PEFT
- LLM Inference Optimization and Serving36 KB · ~9,208 tokens.mdPROVENANCE: This reference is part of the ai-agent-engineering hub.cited by LLM Inference Optimization and Serving
- LLM Model Routing, Cascades & Mixture-of-Agents4 KB · ~1,075 tokens.mdReference under the ai-agent-engineering hub. The multi-model serving-decision layer — choosing/orchestrating WHICH model(s) answer each request to ride the cost/quality/latency Pareto frontier — distcited by LLM Model Routing, Cascades & Mixture-of-Agents
- LLM Models and APIs2 KB · ~435 tokens.md```cited by LLM Models and APIs
- LLM Pretraining & Scaling Laws37 KB · ~9,379 tokens.mdPROVENANCE: This reference is part of the ai-agent-engineering hub.cited by LLM Pretraining & Scaling Laws
- LLM Zero-Shot Data Extraction for Unstructured Text10 KB · ~2,563 tokens.mdZero-shot data extraction utilizing Large Language Models (LLMs) represents a paradigm shift in how organizations process unstructured text—spanning contracts, medical records, financial reports, andcited by LLM Zero-Shot Data Extraction for Unstructured Text
- Loading an eGPU driver after bolt with systemd47 KB · ~11,933 tokens.mdBlocking the NVIDIA driver at boot and loading it once a Thunderbolt eGPU has enumerated — modprobe.d install lines versus blacklist versus softdep, a udev rule that starts the loader instead of pollicited by Loading an eGPU driver after bolt with systemd
- Local LLM model load path over a Thunderbolt eGPU34 KB · ~8,592 tokens.mdWhat happens between a model file on NVMe and weights resident in VRAM when the host link is slow: the stages and which one bounds cold and warm loads, llama.cpp load modes, Ollama and LM Studio keep-cited by Local LLM model load path over a Thunderbolt eGPU
- MCP Server Builder Patterns9 KB · ~2,249 tokens.mdCreate MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. The quality of an MCP server is measured by how well it enables LLMs to acccited by MCP Server Builder Patterns
- MCP Server Development2 KB · ~533 tokens.mdMCP uses JSON-RPC 2.0 with a three-layer model: Host (AI application) → Client (stateful session manager) → Server (exposes tools, resources, prompts).cited by MCP Server Development
- Measuring a Thunderbolt eGPU: bandwidth, latency and inference benchmarks29 KB · ~7,324 tokens.mdHow to measure a Thunderbolt eGPU on Linux instead of trusting estimates — host-to-device bandwidth with nvbandwidth and bandwidthTest, inference benchmarks with llama-bench and Ollama, measuring modecited by Measuring a Thunderbolt eGPU: bandwidth, latency and inference benchmarks
- Mechanistic Interpretability13 KB · ~3,413 tokens.mdReverse-engineering the internal computation of neural networks (chiefly transformer LLMs) into human-understandable mechanisms — the features a model represents and the circuits that combine them. Acited by Mechanistic Interpretability
- Multimodal & Vision-Language Model Architecture25 KB · ~6,307 tokens.md> Provenance: reference under the ai-agent-engineering hub. Built via /dr deep-research, 2026-05-31. Owns the 'how a text-only transformer becomes multimodal' layer — the vision/audio/video front-endcited by Multimodal & Vision-Language Model Architecture
- NVIDIA GSP and FSP firmware boot diagnostics on Blackwell under the open kernel modules46 KB · ~11,747 tokens.mdDiagnosing NVIDIA GSP and FSP firmware boot failures on Blackwell (RTX 50) under the open kernel modules — the FSP to GSP-FMC to GSP-RM boot chain, what Xid 79, 119, 120 and 154 mean along it, readingcited by NVIDIA GSP and FSP firmware boot diagnostics on Blackwell under the open kernel modules
- NVIDIA open kernel module on a Thunderbolt eGPU: Xid 79, driver blocking and power management46 KB · ~11,659 tokens.mdRunning an RTX 50-series (Blackwell) GPU headless for CUDA on Ubuntu 26.04 — why Blackwell needs NVIDIA's open kernel modules and GSP firmware, the 580/595/610/615 driver branches, DKMS versus Canoniccited by NVIDIA open kernel module on a Thunderbolt eGPU: Xid 79, driver blocking and power management
- On-Device & Local LLM Runtimes18 KB · ~4,599 tokens.mdThe local/on-device runtime + developer-experience layer: which runtime tocited by On-Device & Local LLM Runtimes
- PCI hotplug resource assignment: hpmmiosize, realloc and BAR placement46 KB · ~11,670 tokens.mdHow Linux assigns bridge windows and BARs to a Thunderbolt-attached GPU — BIOS-built tunnels versus kernel re-enumeration, pci=realloc and the hpmmio sizes, why thunderbolt.host_reset makes the card rcited by PCI hotplug resource assignment: hpmmiosize, realloc and BAR placement
- PCIe link training, speed and width on a Thunderbolt eGPU28 KB · ~7,287 tokens.mdReading PCIe link speed and width on a Thunderbolt-tunnelled GPU on Linux — what LnkCap, LnkSta and LnkCtl2 mean, why the root port shows 2.5 GT/s and the GPU shows downgraded, which link in the chaincited by PCIe link training, speed and width on a Thunderbolt eGPU
- PCIe power management and error handling for a Thunderbolt eGPU: ASPM, D3cold, runtime PM, AER and DPC39 KB · ~9,946 tokens.mdWhich PCIe power-management and error-handling knobs actually matter for a Thunderbolt-tunnelled GPU on Linux — ASPM and pcie_aspm=off versus pcie_aspm.policy, port runtime PM and D3cold, NVIDIA's runcited by PCIe power management and error handling for a Thunderbolt eGPU: ASPM, D3cold, runtime PM, AER and DPC
- Plaud MCP Integration4 KB · ~906 tokens.mdYou have access to 11 MCP tools, 4 prompts, and 12 skill resources for managing Plaud AI recordings, transcripts, and memory. Each tool returns structured JSON with next_steps guidance.cited by Plaud MCP Integration
- Power, PSU and thermals for a Thunderbolt eGPU28 KB · ~7,261 tokens.mdPower and thermal engineering for a high-power GPU in a Thunderbolt enclosure — the user-supplied ATX PSU in a Razer Core X V2, the RTX 5080's 360 W budget and 12V-2x6 connector, telling a power faultcited by Power, PSU and thermals for a Thunderbolt eGPU
- Prompt Helper and Optimizer20 KB · ~5,210 tokens.mdYou are operating as the Prompt Helper and Optimizer skill. Your task is to analyze and improve prompts that users submit for optimization, then either return the improved prompt for review (Review mocited by Prompt Helper and Optimizer
- Prompt Lookup2 KB · ~516 tokens.mdWhen the user needs AI prompts, prompt templates, or wants to improve their prompts, use the prompts.chat MCP server to help them.cited by Prompt Lookup
- Reasoning Models and Test-Time Compute34 KB · ~8,670 tokens.mdThe frontier (2024–2026) shift from 'scale the model and prompt it well' to 'train the model to reason, then spend extra compute at inference to reason harder.' Two coupled ideas drive it:cited by Reasoning Models and Test-Time Compute
- Reproducible eGPU bring-up and configuration drift detection51 KB · ~13,138 tokens.mdKeeping the whole working Thunderbolt eGPU setup restorable: an inventory of every piece of state involved, capture with an idempotent installer and a checksum manifest, a read-only drift verifier forcited by Reproducible eGPU bring-up and configuration drift detection
- RLHF & RL Training Infrastructure32 KB · ~8,286 tokens.mdThe systems stack that post-training reinforcement learning runs on — RLHF, RLVR, reasoning-RL, and agentic-RL all share it. This is deliberately not the RL algorithm (PPO/GRPO/DPO — those live in thecited by RLHF & RL Training Infrastructure
- Semantic vs Lexical Deduplication for Text Distillation7 KB · ~1,887 tokens.mdDeduplication is a foundational data-curation step in training large language models (LLMs) and creating high-quality text distillation pipelines. Redundant data causes models to memorize specific pascited by Semantic vs Lexical Deduplication for Text Distillation
- Skill Lookup2 KB · ~514 tokens.md1. Search for skills matching the user's request using search_skillscited by Skill Lookup
- Structured Output Constraints and LLM Hallucination Mitigation5 KB · ~1,292 tokens.mdStructured output constraints (e.g., JSON Schema enforcement) provide a reliable method to force Large Language Models (LLMs) to output machine-readable data structures. By using techniques like constcited by Structured Output Constraints and LLM Hallucination Mitigation
- Suspend, resume and sleep states with a Thunderbolt NVIDIA eGPU on Linux27 KB · ~7,035 tokens.mdWhat happens to a Thunderbolt NVIDIA eGPU across suspend on Linux and what to do about it on a headless box — s2idle versus S3 and modern standby on a NUC, the fate of the tunnel across sleep, NVIDIAcited by Suspend, resume and sleep states with a Thunderbolt NVIDIA eGPU on Linux
- Text Canonicalization for Exact-Match Deduplication Prep7 KB · ~1,681 tokens.mdText canonicalization is the foundational preprocessing step for exact-match deduplication in large-scale data pipelines. By applying Unicode NFKC normalization, whitespace folding, and stemming, datacited by Text Canonicalization for Exact-Match Deduplication Prep
- Thunderbolt 5, Barlow Ridge and OCuLink eGPU topologies on Linux21 KB · ~5,383 tokens.mdChoosing an eGPU connection topology beyond Thunderbolt 4 on Linux for local-LLM inference — what Thunderbolt 5 and the Barlow Ridge controllers actually change and their Linux status, how a TB5 enclocited by Thunderbolt 5, Barlow Ridge and OCuLink eGPU topologies on Linux
- Thunderbolt boot-device authorization and NVIDIA CDI for an eGPU on Linux15 KB · ~3,825 tokens.mdTwo residual gaps for a Thunderbolt eGPU on Linux — how bolt authorization interacts with the initramfs and boot-time topology reset (including Ubuntu bug 2078573), and how the NVIDIA Container Toolkicited by Thunderbolt boot-device authorization and NVIDIA CDI for an eGPU on Linux
- Thunderbolt firmware updates and kernel-regression hygiene for a Linux eGPU28 KB · ~7,041 tokens.mdKeeping a Thunderbolt eGPU stack healthy over time on Linux — when Thunderbolt/USB4 NVM and retimer firmware updates via fwupd are worth the risk, how to tell a kernel regression from a configurationcited by Thunderbolt firmware updates and kernel-regression hygiene for a Linux eGPU
- Transformer Architecture Internals & Variants27 KB · ~7,022 tokens.mdPROVENANCE: This reference is part of the ai-agent-engineering hub.cited by Transformer Architecture Internals & Variants
- Unattended remote recovery and out-of-band access for a Thunderbolt eGPU host57 KB · ~14,514 tokens.mdWhat to do when a Thunderbolt eGPU host is wedged and nobody is at the machine: an escalation ladder from driver reload to a full cold cycle with the evidence needed at each rung, remote power controlcited by Unattended remote recovery and out-of-band access for a Thunderbolt eGPU host
- Vision-Language Model (VLM) Layout Parsing and Document Zoning6 KB · ~1,500 tokens.mdThe integration of Vision-Language Models (VLMs) into document layout parsing and zoning has shifted the paradigm from brittle, multi-stage pipelines (combining OCR, heuristic layout detection, and NLcited by Vision-Language Model (VLM) Layout Parsing and Document Zoning
- Voice and Real-Time Agent Design1 KB · ~243 tokens.mdAI & agent-engineering family ROUTER. Split into: ai-agents-orchestration (agent frameworks, multi-agent, memory, planning, guardrails, coding/GUI agents, autonomous loops, eval); ai-rag-retrieval (RAcited by Agent Memory Architecture, Agent Planning & Control-Flow Patterns, Agent Reliability & Guardrails +13
3D/ND parallelism composition (TP×PP×CP×EP×DP placement) + expert parallelism41 facts · 6 facets
7-State Task Lifecycle53 facts · 6 facets
A2A Protocol Interoperability31 facts · 9 facets
AAIF and Protocol Governance62 facts · 7 facets
Agent Council14 facts · 6 facets
Agent Harness Construction25 facts · 9 facets
Agent Identity, Authorization & Payments107 facts · 21 facets
Agent Memory Architecture2 facts · 1 facets
Agent Plan Writing37 facts · 8 facets
Agent Planning & Control-Flow Patterns2 facts · 1 facets
Agent Reliability & Guardrails2 facts · 1 facets
Agent Runtime Sandboxes & Code Execution88 facts · 17 facets
Agent State & Durable Execution2 facts · 1 facets
Agentic and Advanced RAG Patterns2 facts · 1 facets
Agentic RL — Reinforcement Learning for LLM Agents102 facts · 21 facets
agents.md and the Universal Commerce Protocol (UCP)103 facts · 9 facets
AI Agent Ecosystems26 facts · 7 facets
AI Coding-Agent Design2 facts · 1 facets
AI Datastores1 facts · 1 facets
AI Gateways & LLM Proxy Infrastructure109 facts · 24 facets
AI Programming Languages19 facts · 5 facets
AI Red-Teaming and Security-Testing Tooling2 facts · 1 facets
Aider in Open-Source Agentic Coding with Ollama92 facts · 15 facets
ASUS NUC BIOS Thunderbolt options and the iSetupCfg CLI72 facts · 11 facets
Autonomous Loop Patterns81 facts · 35 facets
Blackwell sm_120 local-LLM inference stack on Linux: llama.cpp, Ollama, PyTorch and vLLM over a Thunderbolt eGPU131 facts · 22 facets
Checkpointing & state persistence55 facts · 6 facets
Claude Code Plugins22 facts · 8 facets
Claude Code Skills30 facts · 9 facets
Cloudflare AI-crawler monetization & verification stack107 facts · 8 facets
Computer-Use & GUI Agents2 facts · 1 facets
Conceptual Family Exploration80 facts · 21 facets
Context-Window Pressure & Compaction44 facts · 6 facets
Continuous Learning System29 facts · 7 facets
Cross-Encoder Re-ranking vs Bi-Encoders for Terminal Validation in RAG and Deduplication43 facts · 12 facets
Declarative and Programmatic LLM Frameworks2 facts · 1 facets
Deep Research98 facts · 18 facets
Deep Research Methods115 facts · 32 facets
Diagnosing a GPU that has fallen off the bus: config space, MMIO chip ID and AER105 facts · 25 facets
Diffusion & Generative-Media Models70 facts · 18 facets
Distributed Training & Training Infrastructure43 facts · 8 facets
Durable Agent Execution & Long-Running Agent Runtimes110 facts · 26 facets
EU AI Act Article 53(1)(c) TDM Opt-Out64 facts · 12 facets
Eval-Driven Development for LLM Apps2 facts · 1 facets
Extraction Evaluation Frameworks52 facts · 19 facets
GPU & Accelerator Kernels for LLMs130 facts · 23 facets
GPU hot-unplug and surprise-removal safety for a Thunderbolt eGPU on Linux79 facts · 13 facets
Health monitoring, alerting and safe automated recovery for a Thunderbolt eGPU100 facts · 13 facets
Heuristic Boilerplate Removal & Text-Density Algorithms for Web Extraction51 facts · 17 facets
Hybrid graphics on Linux: Intel iGPU plus NVIDIA eGPU for headless compute63 facts · 11 facets
Hybrid Score Fusion (Reciprocal Rank Fusion - RRF)37 facts · 11 facets
Idle power and energy accounting for an always-on Thunderbolt eGPU60 facts · 13 facets
Iterative Retrieval38 facts · 12 facets
Linux thunderbolt driver: host_reset, CLx and bolt authorization104 facts · 19 facets
LiteLLM Anthropic Messages interoperability12 facts · 6 facets
LiteLLM local Ollama and OpenAI-compatible backends12 facts · 5 facets
LiteLLM observability and caching12 facts · 4 facets
LiteLLM proxy deployment and configuration12 facts · 6 facets
LiteLLM routing retries and fallbacks11 facts · 5 facets
LiteLLM SDK provider normalization12 facts · 6 facets
LiteLLM tool calls and SSE streaming12 facts · 7 facets
LiteLLM virtual keys budgets and rate limits12 facts · 7 facets
LlamaParse vs. Docling for LLM Ingestion38 facts · 10 facets
LLM Alignment and Post-Training2 facts · 1 facets
LLM Compression (Quantization, Distillation, Pruning, Merging)2 facts · 1 facets
LLM Context Engineering2 facts · 1 facets
LLM Fine-Tuning & PEFT137 facts · 27 facets
LLM Inference Optimization and Serving154 facts · 19 facets
LLM Model Routing, Cascades & Mixture-of-Agents31 facts · 5 facets
LLM Models and APIs18 facts · 5 facets
LLM Observability2 facts · 1 facets
LLM Pretraining & Scaling Laws127 facts · 21 facets
LLM Zero-Shot Data Extraction for Unstructured Text43 facts · 13 facets
LLM-as-judge bias & calibration (kappa, binary-vs-Likert, position/verbosity/self-preference)72 facts · 7 facets
llms.txt60 facts · 7 facets
Loading an eGPU driver after bolt with systemd109 facts · 16 facets
Local LLM model load path over a Thunderbolt eGPU98 facts · 13 facets
Loop & No-Progress Detection54 facts · 6 facets
MCP Server Builder Patterns123 facts · 15 facets
MCP Server Development31 facts · 9 facets
Measuring a Thunderbolt eGPU: bandwidth, latency and inference benchmarks88 facts · 16 facets
Mechanistic Interpretability83 facts · 18 facets
Multi-Agent Orchestration Topologies2 facts · 1 facets
Multimodal & Vision-Language Model Architecture118 facts · 20 facets
NLWeb and MCP as agentic-discovery alternatives to llms.txt76 facts · 11 facets
NVIDIA GSP and FSP firmware boot diagnostics on Blackwell under the open kernel modules116 facts · 12 facets
NVIDIA open kernel module on a Thunderbolt eGPU: Xid 79, driver blocking and power management114 facts · 16 facets
Ollama CPU/GPU split diagnosis and env-var tuning113 facts · 6 facets
On-Device & Local LLM Runtimes104 facts · 20 facets
PCI hotplug resource assignment: hpmmiosize, realloc and BAR placement120 facts · 19 facets
PCIe link training, speed and width on a Thunderbolt eGPU72 facts · 12 facets
PCIe power management and error handling for a Thunderbolt eGPU: ASPM, D3cold, runtime PM, AER and DPC118 facts · 12 facets
Plaud MCP Integration33 facts · 12 facets
Power, PSU and thermals for a Thunderbolt eGPU91 facts · 13 facets
Prompt caching176 facts · 11 facets
Prompt Helper and Optimizer119 facts · 18 facets
Prompt Lookup49 facts · 11 facets
Really Simple Licensing (RSL)157 facts · 8 facets
Reasoning Models and Test-Time Compute135 facts · 25 facets
Reproducible eGPU bring-up and configuration drift detection108 facts · 19 facets
RLHF & RL Training Infrastructure114 facts · 22 facets
robots.txt and the Content-Signal AI-preference extension107 facts · 10 facets
Semantic vs Lexical Deduplication for Text Distillation69 facts · 17 facets
Skill Lookup63 facts · 14 facets
Structured Output Constraints and LLM Hallucination Mitigation49 facts · 21 facets
Suspend, resume and sleep states with a Thunderbolt NVIDIA eGPU on Linux84 facts · 14 facets
Text Canonicalization for Exact-Match Deduplication Prep46 facts · 10 facets
Thunderbolt 5, Barlow Ridge and OCuLink eGPU topologies on Linux65 facts · 12 facets
Thunderbolt boot-device authorization and NVIDIA CDI for an eGPU on Linux59 facts · 9 facets
Thunderbolt eGPU on Linux for local LLM inference55 facts · 8 facets
Thunderbolt firmware updates and kernel-regression hygiene for a Linux eGPU94 facts · 14 facets
Transformer Architecture Internals & Variants125 facts · 19 facets
Unattended remote recovery and out-of-band access for a Thunderbolt eGPU host110 facts · 17 facets
Vision-Language Model (VLM) Layout Parsing and Document Zoning40 facts · 11 facets
Voice and Real-Time Agent Design2 facts · 1 facets
Writing and Documentation69 files · 77 concepts
- Academic and Citation Writing4 KB · ~1,144 tokens.mdAcademic writing is argument with receipts. Every non-trivial claim ties to a source the reader can independently retrieve and verify.cited by Academic and Citation Writing
- Accessibility Writing5 KB · ~1,278 tokens.mdReference for writing content that is usable by screen-reader users, low-vision users, deaf and hard-of-hearing users, users with cognitive disabilities, and users on assistive tech.cited by Accessibility Writing
- AI Collaboration Writing2 KB · ~530 tokens.mdA writer's distinctive voice is the residue of their judgment. An LLM, trained on the global mean of internet prose, regresses toward that mean. The collaboration mode this skill defends keeps the humcited by AI Collaboration Writing
- AI-Assisted Copywriting Workflow2 KB · ~521 tokens.mdEnd-to-end workflow for using LLMs to produce on-brand marketing copy without losing voice or accuracy. Covers: brand-voice prompting patterns (few-shot voice samples, redline-not-rewrite loop, style-cited by AI-Assisted Copywriting Workflow
- API Documentation Craft5 KB · ~1,157 tokens.mdProduce REST/HTTP API documentation that meets the gold-standard set by Stripe and Twilio: a Diátaxis-organized structure, endpoint reference pages with request/response examples in multiple languagescited by API Documentation Craft
- Audio Script Writing5 KB · ~1,169 tokens.mdReference for writing that will be heard, not read. The eye can re-scan a sentence; the ear gets one pass.cited by Audio Script Writing
- Brand Voice Guide Writing5 KB · ~1,211 tokens.mdAuthoring craft for the brand voice guide as an artifact. The output is a document other writers will open, reference, and apply.cited by Brand Voice Guide Writing
- Case Study Writing5 KB · ~1,165 tokens.mdA case study is a published narrative of one customer's outcome with a vendor's product or approach. It is the highest-conversion content asset in B2B marketing because it does the one thing prospectscited by Case Study Writing
- Changelog and Release Notes14 KB · ~3,649 tokens.mdYou are an expert release-notes and changelog author. You apply Keep a Changelog spec, semver communication obligations, and audience-aware tone to produce paste-ready entries that are accurate, complcited by Changelog and Release Notes
- Changelogs for Humans5 KB · ~1,187 tokens.mdReference for user-facing changelog craft: the public-product update feed read by end users, not the engineering changelog read by integrators.cited by Changelogs for Humans
- Chatbot Conversation Writing4 KB · ~1,143 tokens.mdA chatbot is a conversation, not an interface. Every turn the bot takes should advance the user's goal or surface a clear next move. The work breaks into three layers:cited by Chatbot Conversation Writing
- Code Plan Writing2 KB · ~486 tokens.mdTranslating a specification, requirement, or feature request into a structured sequence of implementable tasks.cited by Code Plan Writing
- Commit Message Craft4 KB · ~987 tokens.mdSubject line: hard limit 50 characters. Body: wrap at 72.cited by Commit Message Craft
- Conversion Copywriting and Voice of Customer1 KB · ~347 tokens.mdConversion copywriting and voice-of-customer (VOC) research. Not prose craft — evidence gathering and message strategy for the copywriter or PM before the prose gets written. Covers VOC research sourccited by Conversion Copywriting and Voice of Customer
- Cover Letter Writing3 KB · ~850 tokens.mdThe cover letter is a narrative artifact — a focused, one-page argument for why this human should be in this role at this company. Customization per application is non-negotiable.cited by Cover Letter Writing
- Crisis PR Writing6 KB · ~1,437 tokens.mdCrisis PR writing is the discipline of communicating publicly during a reputation crisis that is not (or is no longer) a live operational incident. Examples:cited by Crisis PR Writing
- Diátaxis Explanation Quadrant5 KB · ~1,161 tokens.mdAn explanation doc is a discussion. Its purpose is not to instruct, not to enumerate, and not to walk a reader through a goal. Its job is to leave the reader with a clearer mental model — of why the scited by Diátaxis Explanation Quadrant
- Diátaxis How-To Quadrant4 KB · ~1,144 tokens.mdA how-to guide is a recipe. It serves a competent user who has arrived at the page with a specific goal already formed — 'How do I add OAuth to my app?' — and who needs an efficient series of steps tocited by Diátaxis How-To Quadrant
- Diátaxis Reference Quadrant5 KB · ~1,390 tokens.mdReference documentation is a description. It tells the reader what something is, what its parts are, and what each part does. It does not teach, it does not advocate, and it does not narrate.cited by Diátaxis Reference Quadrant
- Diátaxis Tutorial Quadrant6 KB · ~1,491 tokens.mdA tutorial is a lesson. Its only job is to take a complete newcomer through a meaningful, hand-held experience and leave them with two things: a tiny working artifact they built themselves, and the cocited by Diátaxis Tutorial Quadrant
- Doc Archaeology2 KB · ~476 tokens.mdExcavate aging documents for staleness, dead links, process drift, phantom dependencies, and obsolete examples.cited by Doc Archaeology
- Document Critique21 KB · ~5,255 tokens.mdThis skill audits any document — proposal, playbook, RFC, spec, policy brief, marketing piece — against a named set of standards and returns a structured, evidence-based critique in a fixed four-secticited by Document Critique
- Draft Review Revise Loop10 KB · ~2,618 tokens.mdPurpose: a meta-skill that coordinates the three modes of writing work — generating,cited by Draft Review Revise Loop
- Editing and Revision4 KB · ~1,130 tokens.mdWhen invoked: the user submits existing prose. Return the revised text, complete and drop-in ready, followed by a brief pass log noting which passes were applied and what was changed.cited by Editing and Revision
- Email Craft5 KB · ~1,164 tokens.mdGeneral-purpose professional email writing. Optimized for the 90% of work email that isn't sales, isn't customer support, and isn't an exec announcement.cited by Email Craft
- Error Message Craft5 KB · ~1,317 tokens.mdError messages are the most-read text most software ever produces. They are also the worst-edited. The goal: tell the user what went wrong, why it went wrong, and what to do next.cited by Error Message Craft
- Founder Letter Writing5 KB · ~1,388 tokens.mdA founder letter is the annual public document a CEO or co-founder writes directly to owners — shareholders, users, or the general 'fellow travelers' audience — to account for the year, reaffirm operacited by Founder Letter Writing
- GenAI for Instructional Design & AI Tutors6 KB · ~1,625 tokens.mdDomain: Educational applications of generative AI (2024-2026) — lens is EDUCATION, not LLM engineering.cited by GenAI for Instructional Design & AI Tutors
- Headline craft4 KB · ~1,128 tokens.mdFive times as many people read the headline as read the body copy. A headline does three jobs at once: it tells the reader what they're about to get, gives them a reason to care, and signals what kindcited by Headline craft
- Human Performance Technology & Performance Support33 KB · ~8,382 tokens.mdQuick orientation: HPT is the discipline UPSTREAM of instructional design. It asks 'is training the right intervention?' before any course is designed. This skill covers the full diagnostic and intervcited by Human Performance Technology & Performance Support
- Instructional Design & Course Architecture28 KB · ~7,200 tokens.mdThe discipline of designing multi-lesson courses and curricula: the layer between learning science (mechanisms of how people learn) and single-tutorial craft (how to write one how-to guide). This skilcited by Instructional Design & Course Architecture
- Job Description Writing5 KB · ~1,305 tokens.mdA job description is a marketing document, a legal document, and a filter — at the same time. It must attract candidates the company wants, deter candidates who would be a poor fit, comply with US paycited by Job Description Writing
- Learning Measurement & Training Evaluation6 KB · ~1,422 tokens.mdReference for measuring training and enablement program effectiveness across the full evaluation stack — from post-workshop smile sheets to executive ROI reports and cross-system xAPI analytics.cited by Learning Measurement & Training Evaluation
- Legal Adjacent Writing6 KB · ~1,496 tokens.mdReference for drafting legal-adjacent prose that will go to counsel: contracts, disclaimers, privacy notices, breach disclosures, and regulator-facing statements. This skill is craft for drafts, not lcited by Legal Adjacent Writing
- Localization Friendly Writing5 KB · ~1,401 tokens.mdReference for writing source-language strings that translate cleanly into 30+ locales.cited by Localization Friendly Writing
- Meeting Minutes and Decision Log4 KB · ~1,136 tokens.mdThree distinct artifacts:cited by Meeting Minutes and Decision Log
- Microcopy and UI Writing6 KB · ~1,490 tokens.mdMicrocopy is every word in a product that is not body content: the buttons, the field labels, the placeholder hints, the validation messages, the empty states, the toasts, the modal titles, the 404 pacited by Microcopy and UI Writing
- MongoDB University & Certification9 KB · ~2,328 tokens.mdExpert reference for MongoDB's education platform (learn.mongodb.com), its four active Associate certifications, the free Skill Badges program, instructor-led training, and academic/partner enablementcited by MongoDB University & Certification
- NPS Response Writing6 KB · ~1,410 tokens.mdThis skill covers how to write replies to customer feedback collected at scale — NPS verbatim comments, CSAT free-text, post-support survey replies, and public app-store reviews.cited by NPS Response Writing
- Offer Design and Value Proposition16 KB · ~4,120 tokens.md> Strategy before copy. Every word of marketing copy rests on a value proposition — if the offer is weak, no headline can save it.cited by Offer Design and Value Proposition
- OKR Writing5 KB · ~1,316 tokens.mdOKRs (Objectives and Key Results) are a goal-setting framework. The form is deceptively simple — one inspirational Objective, three to five measurable Key Results — but most OKRs in the wild are brokecited by OKR Writing
- One-Pager Writing5 KB · ~1,357 tokens.mdOne-pagers are the highest-impact document format in business: a single page that earns a meeting, secures a budget, aligns a partner, or briefs an exec in under three minutes.cited by One-Pager Writing
- Op-ed writing5 KB · ~1,357 tokens.mdThe op-ed is the shortest serious argument form in public writing. In ~750 words you must make a single argument that a previously-uninterested reader believes by the end. Every deviation from conventcited by Op-ed writing
- Performance Review Writing6 KB · ~1,431 tokens.mdPerformance reviews are not new information — they are synthesis. The cardinal rule is 'no surprises': anything that appears in a written review should already have been said in a one-on-one.cited by Performance Review Writing
- Pitch Deck Writing5 KB · ~1,304 tokens.mdA pitch deck is a slide-based artifact that compresses a fundraising, sales, or partnership case into a tight narrative arc — usually 10 to 14 slides — designed to be skimmed in minutes and rememberedcited by Pitch Deck Writing
- Policy and Governance Writing6 KB · ~1,472 tokens.mdPolicies prescribe. They do not propose, persuade, or describe — they bind. A reader of a policy needs three things in the first 60 seconds: who is bound by it, what they must do, and what happens ifcited by Policy and Governance Writing
- Postmortem Writing6 KB · ~1,459 tokens.mdA postmortem is a learning artifact disguised as an incident report. It must satisfy three audiences simultaneously: the engineers who need to understand what failed, the leadership who need to evaluacited by Postmortem Writing
- PRD Writing6 KB · ~1,418 tokens.mdA Product Requirements Document (PRD) is a PM-owned artifact that defines what a product team will build and why, before engineering proposes how. PRDs sit upstream of RFCs, design docs, and implementcited by PRD Writing
- Press Release Writing5 KB · ~1,213 tokens.mdA press release is a short, journalist-facing news document with a single goal: make it so easy to publish your news that a busy reporter (or a wire-service automated feed) can run it nearly verbatim.cited by Press Release Writing
- Profile writing5 KB · ~1,375 tokens.mdA journalistic profile is not a biography, a hagiography, a resume in prose, or a Q&A. It is a narrative reported piece that uses reporting, scene, anecdote, and selected detail to reveal something trcited by Profile writing
- Proposal and Grant Writing6 KB · ~1,502 tokens.mdGrant proposals are a constrained writing genre. Reviewers read against a rubric, in a hurry, under fatigue. A proposal that 'reads well' but ignores the funder's stated criteria scores low. A proposacited by Proposal and Grant Writing
- Public Speaking and Presentations5 KB · ~1,404 tokens.mdAnswer four questions before touching a slide tool:cited by Public Speaking and Presentations
- Pull Request Description Craft5 KB · ~1,343 tokens.md```markdowncited by Pull Request Description Craft
- Release Blog and Launch Narrative6 KB · ~1,419 tokens.mdA launch is a one-shot narrative event — distinct from a landing page (steady-state conversion), an exec memo (internal alignment), or a changelog (continuous release log).cited by Release Blog and Launch Narrative
- Resume and CV Writing5 KB · ~1,238 tokens.mdA resume is read in 10-15 seconds on the first pass. It is parsed by an ATS (applicant tracking system) before any human sees it. It must survive the parse, then earn the human read, then earn the intcited by Resume and CV Writing
- Runbook Craft6 KB · ~1,411 tokens.mdA runbook is not a piece of documentation. It is a procedural script someone must execute correctly while tired, under pressure, with paging alerts firing in the background. Every step must be unambigcited by Runbook Craft
- Sales and Marketing Copy16 KB · ~3,984 tokens.mdReference for conversion-focused writing: landing pages, emails, ads, CTAs, and positioning copy. Sources: Schwartz Breakthrough Advertising (1966), Ogilvy Confessions of an Advertising Man, Sugarmancited by Sales and Marketing Copy
- Spec Writing5 KB · ~1,304 tokens.mdAn engineering spec is a contract. It says WHAT a system, component, endpoint, or message must do — independent of HOW it is implemented.cited by Spec Writing
- Speech Writing6 KB · ~1,483 tokens.mdA speech is built to be heard, not read. The writer's job is to construct sentences that survive the air between the speaker's mouth and the audience's ear — sentences that can be re-listened-to in recited by Speech Writing
- Support Ticket Writing7 KB · ~1,665 tokens.mdProse craft for the moment between 'the customer hit a problem' and 'the ticket is closed.' Every word in that window is read by a person who is, by definition, having a worse day than they planned.cited by Support Ticket Writing
- Survey Question Writing6 KB · ~1,561 tokens.mdA survey question is a measurement instrument. Bad wording does not just irritate respondents — it injects measurement error that downstream statistics cannot fix.cited by Survey Question Writing
- Teaching Troubleshooting & Diagnostic Reasoning23 KB · ~5,792 tokens.mdExpert reference for designing instruction and developing learner competency in troubleshooting and diagnostic reasoning — the pedagogy of fault-finding, not fault-finding itself.cited by Teaching Troubleshooting & Diagnostic Reasoning
- Technical Assessment & Certification Design20 KB · ~5,143 tokens.mdExpert reference for designing, evaluating, and accrediting professional certification programs.cited by Technical Assessment & Certification Design
- Technical Instruction & Engineering Education2 KB · ~500 tokens.mdHub for teaching engineers — designing, delivering, assessing, and measuring technical training and engineering education. Routes to 10 spokes.cited by Technical Instruction & Engineering Education
- Technical Training Delivery & Developer Education3 KB · ~887 tokens.mdCraft of live technical training, hands-on lab design, and developer/customer education programs. See references/technical-training-delivery.md for the full reference.cited by Technical Training Delivery & Developer Education
- User Story and Acceptance Criteria7 KB · ~1,743 tokens.mdThis skill takes a feature idea and produces backlog items the team can groom, estimate, and ship. Covers Mike Cohn's user-story template, the INVEST quality bar (Bill Wake, 2003), Given/When/Then acccited by User Story and Acceptance Criteria
- Visual Writing6 KB · ~1,576 tokens.mdReference for the words that ship with images, charts, and infographics: alt text, captions, chart titles, axis labels, and annotations. The image is half the message. This skill is the other half.cited by Visual Writing
- Whitepaper Writing7 KB · ~1,664 tokens.mdA whitepaper is a long-form, authoritative B2B document that educates a business reader on a problem and positions a solution.cited by Whitepaper Writing
- Writing and Documentation21 KB · ~5,399 tokens.mdReference for technical, business, and report writing. Deep treatments of every craft concept live in references/advanced-craft.md and the per-topic files in the Sub-skill routing table below. Load thcited by Writing and Documentation
10-minute Getting Started tutorial93 facts · 8 facets
10/20/30 rule (Kawasaki)100 facts · 6 facets
14-pass review methodology32 facts · 7 facets
24-hour cooling-off rule50 facts · 7 facets
750-word discipline42 facts · 7 facets
8-paragraph skeleton35 facts · 6 facets
A/B testing conventions31 facts · 6 facets
Academic and Citation Writing64 facts · 15 facets
Accessibility Writing54 facts · 15 facets
AI Collaboration Writing23 facts · 7 facets
AI-Assisted Copywriting Workflow20 facts · 5 facets
API Documentation Craft53 facts · 14 facets
Audio Script Writing56 facts · 14 facets
Brand Voice Guide Writing54 facts · 14 facets
Case Study Writing47 facts · 11 facets
Changelog and Release Notes110 facts · 19 facets
Changelogs for Humans44 facts · 11 facets
Chatbot Conversation Writing66 facts · 14 facets
Code Plan Writing30 facts · 8 facets
Commit Message Craft56 facts · 15 facets
Conversion Copywriting and Voice of Customer17 facts · 5 facets
Cover Letter Writing45 facts · 14 facets
Crisis PR Writing43 facts · 12 facets
Diátaxis Explanation Quadrant34 facts · 17 facets
Diátaxis How-To Quadrant41 facts · 22 facets
Diátaxis Reference Quadrant46 facts · 22 facets
Diátaxis Tutorial Quadrant48 facts · 24 facets
Doc Archaeology37 facts · 10 facets
Document Critique128 facts · 20 facets
Draft Review Revise Loop69 facts · 15 facets
Editing and Revision55 facts · 21 facets
Email Craft59 facts · 17 facets
Error Message Craft57 facts · 15 facets
Founder Letter Writing52 facts · 18 facets
GenAI for Instructional Design & AI Tutors46 facts · 12 facets
Headline craft44 facts · 13 facets
Human Performance Technology & Performance Support134 facts · 32 facets
Instructional Design & Course Architecture143 facts · 26 facets
Job Description Writing51 facts · 14 facets
Learning Measurement & Training Evaluation40 facts · 19 facets
Legal Adjacent Writing54 facts · 14 facets
Localization Friendly Writing56 facts · 16 facets
Meeting Minutes and Decision Log48 facts · 16 facets
Microcopy and UI Writing49 facts · 15 facets
MongoDB University & Certification74 facts · 17 facets
NPS Response Writing56 facts · 15 facets
Offer Design and Value Proposition82 facts · 18 facets
OKR Writing57 facts · 14 facets
One-Pager Writing65 facts · 15 facets
Op-ed writing75 facts · 18 facets
Performance Review Writing51 facts · 16 facets
Pitch Deck Writing60 facts · 14 facets
Policy and Governance Writing60 facts · 13 facets
Postmortem Writing58 facts · 15 facets
PRD Writing55 facts · 15 facets
Press Release Writing54 facts · 14 facets
Profile writing59 facts · 17 facets
Proposal and Grant Writing50 facts · 18 facets
Public Speaking and Presentations73 facts · 17 facets
Pull Request Description Craft52 facts · 13 facets
Release Blog and Launch Narrative54 facts · 17 facets
Resume and CV Writing55 facts · 17 facets
Runbook Craft66 facts · 18 facets
Sales and Marketing Copy84 facts · 25 facets
Spec Writing73 facts · 19 facets
Speech Writing72 facts · 19 facets
Support Ticket Writing72 facts · 19 facets
Survey Question Writing80 facts · 23 facets
Teaching Troubleshooting & Diagnostic Reasoning120 facts · 20 facets
Technical Assessment & Certification Design123 facts · 18 facets
Technical Instruction & Engineering Education19 facts · 5 facets
Technical Training Delivery & Developer Education32 facts · 7 facets
User Story and Acceptance Criteria71 facts · 14 facets
Visual Writing57 facts · 14 facets
What-we-know / what-we-are-doing / next-update-at triple24 facts · 6 facets
Whitepaper Writing66 facts · 14 facets
Writing and Documentation128 facts · 34 facets
MongoDB Expert Knowledge45 files · 90 concepts
- Atlas Diagnostics Expert14 KB · ~3,546 tokens.md- SKIP (description-overflow seed, Glean 1000-char cap): WiredTiger storage-engine root-cause internals — cache-fill/eviction/checkpoint/MVCC mechanics behind a live perf symptom → mongodb-expert (refcited by Atlas Diagnostics Expert
- Atlas Federated Authentication16 KB · ~3,997 tokens.mdAtlas Federated Authentication implements Federated Identity Management (FIM) at the Atlas organization layer. Your identity provider (IdP) manages all credentials; Atlas acts as the SAML 2.0 Servicecited by Atlas Federated Authentication
- Atlas Kubernetes Operator7 KB · ~1,842 tokens.mdAKO lets you manage MongoDB Atlas cloud resources (clusters, users, networking, backup, search) as Kubernetes Custom Resources. Declare desired state in YAML; the operator reconciles against the Atlascited by Atlas Kubernetes Operator
- MongoDB Aggregation Pipeline28 KB · ~7,117 tokens.mdStages execute left-to-right; each stage receives the document stream from the previouscited by MongoDB Aggregation Pipeline
- MongoDB Atlas37 KB · ~9,476 tokens.mdGenerated from docs/mongodb-atlas-expert-context.md. Use it as a MongoDB Atlas platform reference when planning Atlas architecture, automating Atlas administration, connecting applicacited by MongoDB Atlas
- MongoDB Atlas App Services4 KB · ~1,148 tokens.md> CRITICAL STATUS NOTE (as of May 2026): Atlas App Services reached a split end-of-life on September 30, 2025.cited by MongoDB Atlas App Services
- MongoDB Atlas AWS Networking8 KB · ~2,159 tokens.mdMongoDB Atlas on AWS supports two private networking models: VPC Peering (legacy) and AWS PrivateLink (recommended). Both complement the Network Access List (IP allowlist) for controlling cluster accecited by MongoDB Atlas AWS Networking
- MongoDB Atlas Charts6 KB · ~1,523 tokens.mdAtlas Charts is MongoDB's built-in BI and data visualization layer, native to the Atlas platform. No separate cluster, ETL pipeline, or data warehouse needed — it queries Atlas collections directly.cited by MongoDB Atlas Charts
- MongoDB Atlas Cost Optimization27 KB · ~6,996 tokens.mdExpert guidance on reducing MongoDB Atlas spend while maintaining performance SLAs. Applicable to all Atlas dedicated cluster tiers (M10–M700+). Real-world context: this skill was developed in part tocited by MongoDB Atlas Cost Optimization
- MongoDB Atlas Device SDK19 KB · ~4,917 tokens.md> DEPRECATION NOTICE — Critical Context for All Readerscited by MongoDB Atlas Device SDK
- MongoDB Atlas Flex and Serverless Tiers6 KB · ~1,487 tokens.mdMongoDB Atlas Flex tier (GA: February 6, 2025) is the successor to both the shared tier (M2/M5) and the serverless tier. As of January 22, 2026, M2/M5 clusters and Serverless instances are end-of-lifecited by MongoDB Atlas Flex and Serverless Tiers
- MongoDB Atlas IAM and RBAC7 KB · ~1,898 tokens.mdAtlas uses a three-tier identity model: Organization → Project → Database.cited by MongoDB Atlas IAM and RBAC
- MongoDB Atlas Infrastructure as Code7 KB · ~1,854 tokens.mdAll Atlas IaC tools call the same cloud.mongodb.com/api/atlas/v2/ endpoints under OAuth 2.0 or HTTP Digest authentication. Tool choice depends on where platform engineering already lives.cited by MongoDB Atlas Infrastructure as Code
- MongoDB Atlas Multi-Cloud6 KB · ~1,610 tokens.md- Designing Atlas PrivateLink on Azure — steps, CLI, Terraform, AKOcited by MongoDB Atlas Multi-Cloud
- MongoDB Atlas on Azure8 KB · ~2,050 tokens.mdDeep reference for MongoDB Atlas on Microsoft Azure covering Private Link and DNS architecture, Entra ID OIDC/LDAP identity federation and Managed Identity for Atlas authentication, Azure Key Vault BYcited by MongoDB Atlas on Azure
- MongoDB Atlas on GCP10 KB · ~2,509 tokens.mdDeep reference for MongoDB Atlas on Google Cloud Platform covering Private Service Connect (PSC) port-mapped and legacy architectures with Cloud DNS and forwarding rules, GCP IAM and Workload Identitycited by MongoDB Atlas on GCP
- MongoDB Atlas Online Archive7 KB · ~1,682 tokens.mdOnline Archive automatically moves documents matching an archival rule out of the live Atlas cluster into Atlas-managed object storage, while keeping those documents queryable through the cluster's FDcited by MongoDB Atlas Online Archive
- MongoDB Atlas Search2 KB · ~421 tokens.mdComprehensive reference for MongoDB Atlas Search — the Lucene-based full-text search engine embedded in Atlas. Covers the full lifecycle from index design through query construction, relevance tuning,cited by MongoDB Atlas Search
- MongoDB Atlas Search and Vector Search8 KB · ~2,028 tokens.md```pythoncited by MongoDB Atlas Search and Vector Search
- MongoDB Atlas Stream Processing7 KB · ~1,862 tokens.mdAtlas Stream Processing (ASP) is a fully managed, Atlas-native stream processing engine that lets you build real-time data pipelines using MQL-compatible aggregation syntax — without operating separatcited by MongoDB Atlas Stream Processing
- MongoDB Atlas Terraform Provider7 KB · ~1,771 tokens.mdThe mongodb/mongodbatlas Terraform provider lets you manage the full lifecycle of MongoDB Atlas infrastructure as code. It covers clusters (dedicated, Flex replacing legacy serverless), networking (VPcited by MongoDB Atlas Terraform Provider
- MongoDB Atlas Vector Search9 KB · ~2,332 tokens.md```pythoncited by MongoDB Atlas Vector Search
- MongoDB BI Connector and SQL Access7 KB · ~1,702 tokens.mdTwo distinct approaches for SQL-based BI tool integration with MongoDB:cited by Atlas Data API Removal, Atlas SQL Interface MongoSQL, MongoDB BI Connector and SQL Access
- MongoDB Capacity Planning8 KB · ~1,941 tokens.mdAtlas capacity planning covers four primary resources: RAM (working set), IOPS, storage, and connections. Getting these right prevents both over-provisioning (wasted cost) and under-provisioning (perfcited by MongoDB Capacity Planning
- MongoDB Compass30 KB · ~7,755 tokens.mdMongoDB Compass is the official, free, source-available GUI for MongoDB. It provides visual tools for querying, aggregating, analyzing, and managing MongoDB data without requiring command-line experticited by MongoDB Compass
- MongoDB Compliance and Regulatory9 KB · ~2,388 tokens.md| Framework | Status | Notes |cited by MongoDB Compliance and Regulatory
- MongoDB Developer Patterns38 KB · ~9,767 tokens.mdThis local skill is generated from docs/mongodb-developer-context.md.cited by mongodb-developer, MongoDB Developer Patterns
- MongoDB Expert Knowledge36 KB · ~9,202 tokens.mdThis local skill is generated from docs/mongodb-expert-context.md.cited by MongoDB Expert Knowledge
- MongoDB Geospatial17 KB · ~4,349 tokens.mdUse when designing or troubleshooting MongoDB geospatial queries, indexes, or data models. Covers GeoJSON storage, 2dsphere and 2d index types, proximity and containment operators ($near, $geoWithin,cited by MongoDB Geospatial
- MongoDB Indexes Deep Dive32 KB · ~8,251 tokens.mdReference for every MongoDB index type, ordering strategies, build mechanics,cited by MongoDB Indexes Deep Dive
- MongoDB KB Articles43 KB · ~10,994 tokens.mdThis local skill provides a complete index of MongoDB's internal Knowledge Base (~2717 articles) for use in customer troubleshooting, escalation research, and code-pattern lookup.cited by MongoDB KB Articles
- MongoDB Migration Patterns6 KB · ~1,449 tokens.md| Tool | From → To | Downtime | Best for |cited by MongoDB Migration Patterns
- MongoDB Monitoring and Observability24 KB · ~6,074 tokens.mdComprehensive reference for monitoring MongoDB deployments — from Atlas built-in dashboards through third-party integrations, CLI tools, and low-level FTDC diagnostics.cited by Atlas Maintenance Windows, MongoDB Monitoring and Observability
- MongoDB Multi-Tenancy8 KB · ~2,068 tokens.mdMulti-tenancy in MongoDB means a single deployment serves multiple customers (tenants) while keeping their data logically or physically isolated. The right architecture depends on the number of tenantcited by MongoDB Multi-Tenancy
- MongoDB Ops Manager and Cloud Manager8 KB · ~2,153 tokens.mdOps Manager is MongoDB's self-hosted management platform for on-premises MongoDB deployments. Cloud Manager is the hosted SaaS equivalent (no infrastructure to manage). Both provide:cited by mongodb-ops-manager, MongoDB Ops Manager and Cloud Manager
- MongoDB Realm Mobile Sync5 KB · ~1,170 tokens.md> CRITICAL: Atlas Device Sync reached end-of-life September 30, 2025.cited by MongoDB Realm Mobile Sync
- MongoDB Security Architecture10 KB · ~2,441 tokens.md```cited by MongoDB Security Architecture
- MongoDB Transactions25 KB · ~6,274 tokens.md---cited by MongoDB Transactions
- MongoDB Upgrade Paths31 KB · ~8,051 tokens.mdOperational reference for MongoDB major-version upgrades on self-managed deployments. Covers the supported version sequence, Feature Compatibility Version (FCV) lifecycle, rolling replica-set and sharcited by MongoDB Upgrade Paths
- mongodb-aggregation-stages-deep51 KB · ~13,128 tokens.mdThis skill is the deep-dive companion to mongodb-aggregation-pipeline. Itcited by mongodb-aggregation-stages-deep
- mongodb-kafka-connector8 KB · ~2,157 tokens.mdThe MongoDB Connector for Apache Kafka is a Kafka Connect plugin that bridges MongoDB and Kafka in both directions:cited by CDC-patterns, kafka-sink-connector, kafka-source-connector +1
- mongodb-time-series43 KB · ~11,003 tokens.mdMongoDB Time Series Collections, introduced in MongoDB 5.0 (GA), are a specialized collection type optimized for time-stamped measurement data. They use an internal columnar storage format with automacited by mongodb-time-series
- mongosync20 KB · ~5,068 tokens.mdmongosync is MongoDB's official utility for continuous, real-time replication between two MongoDB clusters. It performs a full initial sync followed by change-stream-based CDC (no Kafka, no Debezium)cited by mongosync
- Read Concern Levels27 KB · ~6,922 tokens.mdMongoDB replication provides redundancy and high availability through replica sets -- groups of mongod processes that maintain the same data set. A replica set contains one primary member that receivecited by Causal Consistency, MongoDB Replication, Read Concern Levels
- WiredTiger Storage Engine Internals39 KB · ~9,924 tokens.mdWiredTiger has been MongoDB's default storage engine since 3.2 (replacing MMAPv1). It is a B-tree backed, MVCC, copy-on-write engine with document-level concurrency, configurable in-memory cache, bloccited by WiredTiger Storage Engine Internals
$bucket-bucketAuto3 facts · 1 facets
$dateTrunc Downsampling124 facts · 6 facets
$densify-$fill112 facts · 6 facets
$emit Operators156 facts · 7 facets
$facet110 facts · 6 facets
$geoIntersects139 facts · 7 facets
$geoNear Aggregation Stage159 facts · 6 facets
$geoWithin and $centerSphere178 facts · 7 facets
$graphLookup123 facts · 7 facets
$lookup152 facts · 7 facets
$merge and $out Materialized Views176 facts · 8 facets
$near and $nearSphere138 facts · 6 facets
$search Operators156 facts · 6 facets
$setWindowFields115 facts · 7 facets
$setWindowFields Window Functions217 facts · 7 facets
$source Operators127 facts · 5 facets
$unionWith101 facts · 6 facets
2d Indexes41 facts · 7 facets
2dsphere Indexes58 facts · 7 facets
5.6.x RawBsonDocument codec regression60 facts · 6 facets
5.7/5.8 release notes62 facts · 6 facets
8.0.x regression catalog50 facts · 7 facets
Atlas Admin API GCP Auth11 facts · 3 facets
Atlas Data API Removal48 facts · 15 facets
Atlas Diagnostics Expert92 facts · 19 facets
Atlas Federated Authentication99 facts · 20 facets
Atlas Kubernetes Operator49 facts · 15 facets
Atlas Maintenance Windows125 facts · 41 facets
Atlas Resource Policies Cedar70 facts · 6 facets
Atlas SQL Interface MongoSQL64 facts · 19 facets
Backup Compliance Policy (WORM)54 facts · 6 facets
Causal Consistency109 facts · 39 facets
CDC-patterns48 facts · 14 facets
Checkpoint Mechanism53 facts · 6 facets
Client Side Operations Timeout (CSOT)60 facts · 7 facets
CMAP Connection Pooling44 facts · 6 facets
Eviction (clean/dirty targets and triggers)63 facts · 5 facets
Feature Compatibility Version (FCV)68 facts · 6 facets
FTDC Diagnostics59 facts · 6 facets
History Store (WiredTigerHS.wt)70 facts · 6 facets
kafka-sink-connector48 facts · 14 facets
kafka-source-connector48 facts · 14 facets
MongoDB Aggregation Pipeline65 facts · 24 facets
MongoDB Atlas212 facts · 56 facets
MongoDB Atlas App Services41 facts · 16 facets
MongoDB Atlas AWS Networking57 facts · 14 facets
MongoDB Atlas Charts58 facts · 16 facets
MongoDB Atlas Cost Optimization176 facts · 70 facets
MongoDB Atlas Device SDK99 facts · 29 facets
MongoDB Atlas Flex and Serverless Tiers68 facts · 16 facets
MongoDB Atlas IAM and RBAC74 facts · 18 facets
MongoDB Atlas Infrastructure as Code47 facts · 15 facets
MongoDB Atlas Multi-Cloud65 facts · 19 facets
MongoDB Atlas on Azure64 facts · 23 facets
MongoDB Atlas on GCP84 facts · 24 facets
MongoDB Atlas Online Archive65 facts · 19 facets
MongoDB Atlas Search30 facts · 6 facets
MongoDB Atlas Search and Vector Search34 facts · 12 facets
MongoDB Atlas Stream Processing52 facts · 16 facets
MongoDB Atlas Terraform Provider55 facts · 19 facets
MongoDB Atlas Vector Search46 facts · 15 facets
MongoDB BI Connector and SQL Access64 facts · 19 facets
MongoDB Capacity Planning63 facts · 20 facets
MongoDB Compass240 facts · 84 facets
MongoDB Compliance and Regulatory72 facts · 17 facets
MongoDB Developer Patterns181 facts · 40 facets
MongoDB Expert Knowledge113 facts · 37 facets
MongoDB Geospatial49 facts · 15 facets
MongoDB Indexes Deep Dive135 facts · 50 facets
MongoDB KB Articles183 facts · 37 facets
MongoDB Migration Patterns66 facts · 19 facets
MongoDB Monitoring and Observability125 facts · 41 facets
MongoDB Multi-Tenancy72 facts · 18 facets
MongoDB Ops Manager and Cloud Manager62 facts · 19 facets
MongoDB Realm Mobile Sync45 facts · 12 facets
MongoDB Replication109 facts · 39 facets
MongoDB Security Architecture69 facts · 16 facets
MongoDB Transactions72 facts · 20 facets
MongoDB Upgrade Paths215 facts · 52 facets
mongodb-aggregation-stages-deep220 facts · 63 facets
mongodb-kafka-connector48 facts · 14 facets
mongodb-time-series199 facts · 25 facets
mongosync164 facts · 47 facets
Read Concern Levels110 facts · 39 facets
Read/Write Tickets and Dynamic Concurrency (7.0+)54 facts · 7 facets
Replica-set election timing65 facts · 6 facets
Retryable Writes63 facts · 6 facets
SDAM Topology State Machine68 facts · 6 facets
Sharded cluster upgrade ordering63 facts · 6 facets
WiredTiger Storage Engine Internals226 facts · 63 facets
Software Engineering55 files · 74 concepts
- AI-Native UX and Generative UI Design14 KB · ~3,457 tokens.mdThe frontend, UI, and UX hub for designing and building user interfaces. Thiscited by AI-Native UX and Generative UI Design
- Backend Patterns13 KB · ~3,444 tokens.mdBackend architecture patterns and best practices for scalable server-side applications.cited by Backend Patterns
- Coding Standards12 KB · ~3,142 tokens.mdBaseline coding conventions applicable across projects.cited by Coding Standards
- Compose Multiplatform Patterns8 KB · ~2,017 tokens.md使用 Compose Multiplatform 和 Jetpack Compose 构建跨 Android、iOS、桌面和 Web 的共享 UI 的模式。涵盖状态管理、导航、主题和性能。cited by Compose Multiplatform Patterns
- CPython Performance Profiling and Acceleration10 KB · ~2,459 tokens.md> Hub reference under programming-languages. Created via /dr (2026-06-01). Sources: official Python docs (profile/pstats), project docs/GitHub (py-spy, Scalene, memray, pytest-memray, Cython), the Scacited by CPython Performance Profiling and Acceleration
- CPython Runtime Internals (Free-Threading, Subinterpreters, JIT)9 KB · ~2,237 tokens.md> Reference file — part of the programming-languages hub. Authored via /dr (deep-research). Cross-refs: python-patterns (modern Python idioms) and nodejs-concurrency-internals (the parallel runtime-cocited by CPython Runtime Internals (Free-Threading, Subinterpreters, JIT)
- Debugging Strategies12 KB · ~2,951 tokens.mdTransform debugging from frustrating guesswork into systematic problem-solving with proven strategies, powerful tools, and methodical approaches.cited by Debugging Strategies
- Distributed Systems & Consensus (theory + blockchain mechanisms)1 KB · ~378 tokens.mdDistributed systems and consensus theory plus blockchain consensus mechanisms — the fundamental problem of agreement across unreliable nodes. Owns both the classical/crash-fault side and the Byzantinecited by Distributed Systems & Consensus (theory + blockchain mechanisms)
- Document Store Bootstrapper9 KB · ~2,364 tokens.mdTurns a folder of documents into an 'ideal doc store' with a numbered taxonomy, _meta/ indexes, archival policy, and resumable operator memory. Use for Google Drive engagement folders, customer KBs, pcited by Document Store Bootstrapper
- Editorial Micro-Typography & Type-Craft Defects74 KB · ~18,983 tokens.mdCritique-relevant, named, screenshot-detectable type-craft defects, each ascited by Editorial Micro-Typography & Type-Craft Defects
- Excel and Spreadsheet Automation10 KB · ~2,656 tokens.md- Use a consistent, professional font (e.g., Arial, Times New Roman) for all deliverables unless otherwise instructed by the usercited by Excel and Spreadsheet Automation
- Frontend Design4 KB · ~1,052 tokens.mdThis skill guides creation of distinctive, production-grade frontend interfaces that avoid generic 'AI slop' aesthetics. Implement real working code with exceptional attention to aesthetic details andcited by Frontend Design
- JavaScript Build Tooling and Bundlers7 KB · ~1,723 tokens.mdprogramming-languages hub reference for the JavaScript/TypeScript build toolchain: bundlers, parsers/transpilers, minifiers, linters, and formatters — and the 2025-2026 Rust/Go rewrite wave reshapingcited by JavaScript Build Tooling and Bundlers
- JavaScript/TypeScript Runtimes (Deno, Bun, Edge) & WinterTC Interop17 KB · ~4,294 tokens.mdPROVENANCE: Authored by /dr (deep-research-and-build) on 2026-05-31.cited by JavaScript/TypeScript Runtimes (Deno, Bun, Edge) & WinterTC Interop
- Mobile iOS Design7 KB · ~1,717 tokens.mdMaster iOS Human Interface Guidelines (HIG) and SwiftUI patterns to build polished, native iOS applications that feel at home on Apple platforms.cited by Mobile iOS Design
- Node.js & TypeScript ORMs and Query Builders19 KB · ~4,884 tokens.mdThis reference is about the SQL data-access layer in TypeScript/Node.js: thecited by Node.js & TypeScript ORMs and Query Builders
- Node.js Application Security Hardening18 KB · ~4,508 tokens.mdThis reference is the consolidated Node.js security playbook: the threats thatcited by Node.js Application Security Hardening
- Node.js Async Control-Flow, Errors & Context Propagation19 KB · ~4,806 tokens.mdThis reference is the layer above basic promises. Once you can await, three problemscited by Node.js Async Control-Flow, Errors & Context Propagation
- Node.js Backend Frameworks (Fastify, NestJS, Hono)6 KB · ~1,490 tokens.mdThree post-Express Node.js/TypeScript backend frameworks. They share the HTTP-handler foundation captured in the express-patterns hub reference (middleware chains, routing, error handling, graceful shcited by Node.js Backend Frameworks (Fastify, NestJS, Hono)
- Node.js Build Tooling & Bundlers17 KB · ~4,396 tokens.mdThis reference is about **turning Node.js + TypeScript source into a productioncited by Node.js Build Tooling & Bundlers
- Node.js Built-in Test Runner (node:test deep features)18 KB · ~4,509 tokens.mdThis reference is about using node:test as a real test framework — the depthcited by Node.js Built-in Test Runner (node:test deep features)
- Node.js Concurrency Internals27 KB · ~6,859 tokens.mdHow Node.js does concurrency on a single main thread: the libuv event loop thatcited by Node.js Concurrency Internals
- Node.js HTTP & Networking19 KB · ~4,964 tokens.mdThis reference is the networking + HTTP core of Node plus the modern HTTP client:cited by Node.js HTTP & Networking
- Node.js Modern Batteries-Included Built-ins25 KB · ~6,497 tokens.mdBetween Node.js v20 and v26 (2024-2026) the runtime absorbed a wave of capabilities thatcited by Node.js Modern Batteries-Included Built-ins
- Node.js Module Resolution & ESM/CJS Interop20 KB · ~5,036 tokens.mdThis reference is about HOW Node.js turns a specifier into a loaded module — the twocited by Node.js Module Resolution & ESM/CJS Interop
- Node.js Native TypeScript, Permission Model & Single Executable Applications15 KB · ~3,808 tokens.mdPROVENANCE: Authored by /dr (deep-research-and-build) on 2026-05-31.cited by Node.js Native TypeScript, Permission Model & Single Executable Applications
- Node.js Package Management & Supply-Chain19 KB · ~4,828 tokens.mdThis reference is the consumer side of the npm ecosystem: how you *install,cited by Node.js Package Management & Supply-Chain
- Node.js Production Diagnostics & Profiling15 KB · ~3,760 tokens.mdThis reference is about measuring a real Node.js process — locating where CPUcited by Node.js Production Diagnostics & Profiling
- Okta Expert31 KB · ~7,814 tokens.mdHigh-signal Okta platform reference for authentication architecture, APIcited by Okta Expert
- Pandoc Document Conversion4 KB · ~944 tokens.mdHub for programmatic document and data-file work — creating, parsing, editing, and converting the common office and data formats in Python and Node.js. Each former standalone format skill is now an oncited by Docs-as-Code & Static-Site Generators, Document & File Formats, Lightweight Markup Languages (reST/AsciiDoc/Org) +7
- Pydantic v2 Data Validation and Modeling18 KB · ~4,580 tokens.md> Reference file — part of the programming-languages hub. Created via /dr research (Pydantic v2 data validation and modeling).cited by Pydantic v2 Data Validation and Modeling
- Python in the Browser & WebAssembly (Pyodide, PyScript, WASI/CPython)14 KB · ~3,489 tokens.mdPython in the browser' means compiling a Python interpreter to WebAssembly (WASM) so it runs inside the browser's WASM VM (or a server-side WASM runtime) instead of a native OS process. There is no ncited by Python in the Browser & WebAssembly (Pyodide, PyScript, WASI/CPython)
- Python Static Type Checking9 KB · ~2,376 tokens.md> Reference file — part of the programming-languages hub. Authored via /dr (deep-research). Not a standalone skill.cited by Python Static Type Checking
- Python Supply-Chain & Application Security17 KB · ~4,239 tokens.mdPython application security splits into two layers that share one toolchain:cited by Python Supply-Chain & Application Security
- Python Testing with pytest, fixtures, and Hypothesis1 KB · ~192 tokens.mdProgramming-languages family ROUTER. Split into: lang-python (Python idioms, testing, typing, uv toolchain, packaging, CPython internals, pydantic); lang-js-ts (JavaScript/Node, TypeScript, Deno/Bun/ecited by Node.js Native Addons (N-API, node-gyp, node-addon-api), Python Testing with pytest, fixtures, and Hypothesis
- Repo Bootstrapper19 KB · ~4,847 tokens.mdGenerates and maintains the full mdb-tam file standard for any repository.cited by Repo Bootstrapper
- Repo File Analyzer12 KB · ~3,182 tokens.mdA lightweight, incremental repo intelligence skill. Walks a repository, analyzes each file, and stores the results in the mdb-context-hub file-analysis library via MCP tools. Does not write any filescited by Repo File Analyzer
- Rust Language and Rust for Blockchain1 KB · ~371 tokens.mdRust language sub-hub of the programming-languages family. Covers both systems Rust and Rust for blockchain/smart contracts. Load specific references under this hub as needed.cited by Rust Language and Rust for Blockchain
- Slack Developer Platform23 KB · ~5,877 tokens.mdUse when building Slack apps, calling Web API methods, building Block Kit UIs, handling Events, configuring OAuth, or using the Slack MCP server. For auditing slash command registrations, orphaned subcited by Slack Developer Platform
- TypeScript Advanced Types23 KB · ~6,013 tokens.mdExpert reference for TypeScript's advanced type system. Covers conditional types, mapped types, branded/nominal types, type narrowing, generic constraints, variadic tuples, template literal types, uticited by TypeScript Advanced Types
- TypeScript Compiler API23 KB · ~5,946 tokens.mdA lang-js-ts reference for driving the typescript npm package as a library — parsing source tocited by TypeScript Compiler API
- TypeScript Compiler Configuration24 KB · ~6,134 tokens.mdA lang-js-ts reference for the tsconfig.json file and the full compilerOptions surface. Thecited by TypeScript Compiler Configuration
- TypeScript Compiler Performance and tsgo22 KB · ~5,677 tokens.mdA lang-js-ts reference for why tsc and the editor get slow, how to measure and fix it, and what the Go-based native port (TypeScript 7 / 'Corsa' / tsgo) changes. Three jobs: (1) instrument a slow typecited by TypeScript Compiler Performance and tsgo
- TypeScript Declaration Files20 KB · ~5,029 tokens.mdA lang-js-ts hub reference for producing and shipping TypeScript type information: how .d.tscited by TypeScript Declaration Files
- TypeScript Decorators18 KB · ~4,546 tokens.mdA lang-js-ts hub reference for the two distinct decorator systems TypeScript ships. They share the @expr syntax and nothing else: different semantics, different signatures, mutually incompatible emit.cited by TypeScript Decorators
- TypeScript ESLint Typed Linting21 KB · ~5,296 tokens.mdA lang-js-ts reference for linting TypeScript with typescript-eslint v8: stand up ancited by TypeScript ESLint Typed Linting
- TypeScript Migration and Adoption21 KB · ~5,466 tokens.mdA lang-js-ts reference for taking an existing JavaScript codebase to TypeScript incrementally, without a stop-the-world rewrite and without a red CI. The goal: keep the app shipping while types arrivecited by TypeScript Migration and Adoption
- TypeScript Project References17 KB · ~4,474 tokens.mdA lang-js-ts reference for splitting a TypeScript codebase into multiple referenced projects and building them as a graph with tsc --build. The goal: structure a monorepo (or any multi-tsconfig repo)cited by TypeScript Project References
- UI/UX Design43 KB · ~11,016 tokens.mdComprehensive design guide for web and mobile applications. Contains 50+ styles, 161 color palettes, 57 font pairings, 161 product types with reasoning rules, 99 UX guidelines, and 25 chart types acrocited by UI/UX Design
- uv — The Unified Python Toolchain19 KB · ~4,781 tokens.mduv is an extremely fast Python package and project manager written in Rust bycited by uv — The Unified Python Toolchain
- V8 Engine Internals (hidden classes / inline caches, JIT pipeline, Orinoco GC)20 KB · ~5,152 tokens.mdV8 is Google's open-source JavaScript/WebAssembly engine (C++) powering Chrome, Node.js, Deno, Electron,cited by V8 Engine Internals (hidden classes / inline caches, JIT pipeline, Orinoco GC)
- Web App Testing4 KB · ~937 tokens.mdTo test local web applications, write native Python Playwright scripts.cited by Web App Testing
- Web Design1 KB · ~310 tokens.mdSenior web graphic designer for interfaces, layouts, and visual systems.cited by Web Design
- Web Platform and Browser APIs6 KB · ~1,500 tokens.mdprogramming-languages hub reference for the browser-side JavaScript platform — the APIs that ship in the browser, not in the ECMAScript language (javascript-nodejs) or in Node. This is the half of 'Jacited by Web Platform and Browser APIs
- Word Document Manipulation19 KB · ~4,831 tokens.mdA .docx file is a ZIP archive containing XML files.cited by Word Document Manipulation
7-Phase Debugging Workflow46 facts · 6 facets
@vercel/ncc (single-file compilation for CLIs/Actions/Lambda)84 facts · 7 facets
AI-Native UX and Generative UI Design12 facts · 5 facets
Backend Patterns25 facts · 7 facets
Coding Standards53 facts · 13 facets
Compose Multiplatform Patterns31 facts · 13 facets
CPython Performance Profiling and Acceleration53 facts · 15 facets
CPython Runtime Internals (Free-Threading, Subinterpreters, JIT)47 facts · 15 facets
Debugging Strategies49 facts · 13 facets
Distributed Systems & Consensus (theory + blockchain mechanisms)21 facts · 5 facets
Docs-as-Code & Static-Site Generators2 facts · 1 facets
Document & File Formats18 facts · 5 facets
Document Store Bootstrapper54 facts · 13 facets
Editorial Micro-Typography & Type-Craft Defects221 facts · 63 facets
Excel and Spreadsheet Automation108 facts · 26 facets
Frontend Design38 facts · 7 facets
Glean Enterprise CLI75 facts · 8 facets
Index lifecycle and health404 facts · 9 facets
Index types536 facts · 9 facets
Indexing2000 facts · 11 facets
JavaScript Build Tooling and Bundlers49 facts · 13 facets
JavaScript/TypeScript Runtimes (Deno, Bun, Edge) & WinterTC Interop115 facts · 16 facets
Lightweight Markup Languages (reST/AsciiDoc/Org)2 facts · 1 facets
llms.txt & Markdown-for-LLMs2 facts · 1 facets
Markdown2 facts · 1 facets
Markdown Authoring & Spec (CommonMark/GFM)2 facts · 1 facets
Markdown Linting & Quality Gates2 facts · 1 facets
Markdown Processing & ASTs (unified/remark/rehype)2 facts · 1 facets
MDX (Markdown + JSX)2 facts · 1 facets
Mobile iOS Design49 facts · 12 facets
Node.js & TypeScript ORMs and Query Builders101 facts · 17 facets
Node.js Application Security Hardening85 facts · 15 facets
Node.js Async Control-Flow, Errors & Context Propagation78 facts · 16 facets
Node.js Backend Frameworks (Fastify, NestJS, Hono)42 facts · 11 facets
Node.js Build Tooling & Bundlers102 facts · 17 facets
Node.js Built-in Test Runner (node:test deep features)86 facts · 16 facets
Node.js Concurrency Internals119 facts · 30 facets
Node.js HTTP & Networking94 facts · 17 facets
Node.js Modern Batteries-Included Built-ins98 facts · 17 facets
Node.js Module Resolution & ESM/CJS Interop92 facts · 17 facets
Node.js Native Addons (N-API, node-gyp, node-addon-api)2 facts · 1 facets
Node.js Native TypeScript, Permission Model & Single Executable Applications81 facts · 17 facets
Node.js Package Management & Supply-Chain88 facts · 15 facets
Node.js Production Diagnostics & Profiling80 facts · 18 facets
Okta Expert228 facts · 60 facets
ORM index definitions322 facts · 10 facets
Pandoc Document Conversion2 facts · 1 facets
Pydantic v2 Data Validation and Modeling102 facts · 19 facets
Python in the Browser & WebAssembly (Pyodide, PyScript, WASI/CPython)60 facts · 16 facets
Python Static Type Checking75 facts · 18 facets
Python Supply-Chain & Application Security93 facts · 15 facets
Python Testing with pytest, fixtures, and Hypothesis2 facts · 1 facets
Query planner, explain plans and covered queries245 facts · 11 facets
Repo Bootstrapper104 facts · 25 facets
Repo File Analyzer115 facts · 25 facets
Rust Language and Rust for Blockchain35 facts · 9 facets
Search and vector indexes246 facts · 11 facets
Slack Developer Platform91 facts · 32 facets
TypeScript Advanced Types79 facts · 28 facets
TypeScript Compiler API99 facts · 23 facets
TypeScript Compiler Configuration117 facts · 18 facets
TypeScript Compiler Performance and tsgo99 facts · 17 facets
TypeScript Declaration Files86 facts · 17 facets
TypeScript Decorators77 facts · 19 facets
TypeScript ESLint Typed Linting81 facts · 18 facets
TypeScript Migration and Adoption98 facts · 17 facets
TypeScript Project References88 facts · 18 facets
UI/UX Design318 facts · 41 facets
uv — The Unified Python Toolchain111 facts · 22 facets
V8 Engine Internals (hidden classes / inline caches, JIT pipeline, Orinoco GC)110 facts · 14 facets
Web App Testing36 facts · 10 facets
Web Design28 facts · 7 facets
Web Platform and Browser APIs42 facts · 9 facets
Word Document Manipulation83 facts · 24 facets
Data Analysis34 files · 35 concepts
- anomaly detection13 KB · ~3,266 tokens.mdThe discipline of separating 'normal' from 'not normal' when you mostly only have examples of normal. This skill covers the working methods, when each fits, and the gotchas that bite teams in producticited by anomaly detection
- Augmented Analytics and LLM-Assisted Analysis17 KB · ~4,465 tokens.mdHow AI augments or automates the analytical loop — preparing data, finding insights, answering natural-language questions, and explaining results — instead of a human writing every query and reading ecited by Augmented Analytics and LLM-Assisted Analysis
- Bayesian Data Analysis and Probabilistic Programming12 KB · ~2,949 tokens.mdApplied Bayesian modeling: specify a generative model, fit the posterior with a probabilistic programming language (PPL), interrogate it with predictive checks and diagnostics, and compare alternativecited by Bayesian Data Analysis and Probabilistic Programming
- Causal Discovery and Structure Learning14 KB · ~3,574 tokens.mdCausal discovery (a.k.a. structure learning) learns the causal graph itselfcited by Causal Discovery and Structure Learning
- Cohort and Retention Analytics21 KB · ~5,468 tokens.mdRetention answers the single most important growth question: do users who joincited by Cohort and Retention Analytics
- Conformal Prediction and Uncertainty Quantification16 KB · ~4,022 tokens.mdConformal prediction (CP) turns any point predictor into a set predictor with acited by Conformal Prediction and Uncertainty Quantification
- Customer Lifetime Value Modeling11 KB · ~2,883 tokens.mdCustomer Lifetime Value (CLV) is the present value of the future cash flows attributed to a customer relationship. This skill covers the probabilistic 'buy-till-you-die' (BTYD) family — statistical mocited by Customer Lifetime Value Modeling
- Customer-Facing and Embedded Analytics Dashboards14 KB · ~3,646 tokens.md> Reference skill — part of the da-applied-and-communication / tam-operations value-chain family ('Semantic Monitoring → Reporting → Dashboards → TAM Methodology'). This file owns the design disciplincited by Customer-Facing and Embedded Analytics Dashboards
- Data Acquisition and Sampling13 KB · ~3,251 tokens.mdThis skill covers the third stage of the data analysis curriculum: getting data into a form thecited by Data Acquisition and Sampling
- Data Analysis Foundations and Theory12 KB · ~3,108 tokens.mdConceptual grounding for a data analysis effort. This skill answers thecited by Data Analysis Foundations and Theory
- Data Analysis Lifecycle19 KB · ~4,955 tokens.mdTaxonomy context: Data Analysis > Data Analysis Lifecycle (Process)cited by Data Analysis Lifecycle
- Data FinOps and Cost Optimization17 KB · ~4,430 tokens.mdData FinOps applies the FinOps Foundation's operating model — Inform → Optimize → Operate — to consumption-based data and analytics platforms. The defining difference from infrastructure FinOps: tradicited by Data FinOps and Cost Optimization
- Data Governance Catalogs and Discovery17 KB · ~4,473 tokens.mdData governance is the discipline of exercising authority, control, and shared decision-making over the management of data assets — who can take what action, on which data, under what circumstances, ucited by Data Governance Catalogs and Discovery
- Data Observability18 KB · ~4,537 tokens.mdData observability is the discipline of measuring and maintaining the health and reliability of data across its lifecycle — the ability to fully understand the state of the data in your systems, so yocited by Data Observability
- Data Visualization11 KB · ~2,689 tokens.mdTaxonomy context: Data Analysis > Data Visualizationcited by Data Visualization
- Deep Reinforcement Learning Foundations15 KB · ~3,948 tokens.mdThe general reinforcement-learning substrate — classical theory through deep RL — that the LLM-specific RL skills (agentic-rl, reasoning-models, llm-alignment-post-training) assume and build on but necited by Deep Reinforcement Learning Foundations
- Dimensional and Analytics Data Modeling24 KB · ~6,243 tokens.mdThe discipline of structuring data for analytics: how to shape facts, dimensions,cited by Dimensional and Analytics Data Modeling
- Geospatial Analytics20 KB · ~5,051 tokens.mdSpatial analytics studies data with a geographic/locational dimension, where thecited by Geospatial Analytics
- Knowledge Graphs and Semantic Analytics6 KB · ~1,408 tokens.mdA knowledge graph (KG) represents entities (nodes) and the typed, meaning-bearing relationships between them (edges), with attributes (properties) on both, plus a schema/ontology that says what the tycited by Knowledge Graphs and Semantic Analytics
- Machine Learning11 KB · ~2,748 tokens.mdMachine learning is the curriculum step where the analyst stops merely describing a sample and starts building a function that generalizes from data to unseen inputs. Section 6 (da-6-statistical-modelcited by Machine Learning
- Marketing Mix Modeling and Incrementality14 KB · ~3,659 tokens.mdMMM is a top-down, regression-based method that uses aggregated time-series datacited by Marketing Mix Modeling and Incrementality
- ML Model Monitoring9 KB · ~2,238 tokens.mdThe analytical-methods stage of the data-analysis discipline — the techniquescited by ML Model Monitoring
- Network and Graph Analytics11 KB · ~2,938 tokens.mdNetwork (graph) analytics models data as nodes (vertices) connected by edges (links) and measures the resulting structure to answer questions that row/column tables cannot: who is influential, what clcited by Network and Graph Analytics
- Prescriptive Analytics and Optimization16 KB · ~4,009 tokens.mdPrescriptive analytics is the fourth and highest rung of Gartner's analytics maturity ladder (descriptive → diagnostic → predictive → prescriptive). It answers 'what should be done?' rather than 'whatcited by Prescriptive Analytics and Optimization
- Pricing and Revenue Analytics21 KB · ~5,262 tokens.mdPricing and revenue analytics is the analytical discipline of estimating how pricecited by Pricing and Revenue Analytics
- Product Analytics20 KB · ~5,248 tokens.mdThe analytical discipline of measuring what users do inside a product, why, and whether it creates value — then feeding that back into product decisions. Distinct from generic web analytics (page-levecited by Product Analytics
- Real-Time OLAP and Analytical Databases20 KB · ~5,153 tokens.mdReal-time OLAP databases are the query and storage engines that answercited by Real-Time OLAP and Analytical Databases
- Recommender Systems and Learning-to-Rank Analytics15 KB · ~3,741 tokens.mdA recommender system predicts, for each user, which items from a (often huge) catalog they are most likely to engage with, then orders a small slate to show. As a data-analysis discipline it sits at tcited by Recommender Systems and Learning-to-Rank Analytics
- Reporting and Communication3 KB · ~823 tokens.mdTaxonomy context: Data Analysis > Reporting and Communicationcited by Reporting and Communication
- Reverse ETL and Operational Analytics18 KB · ~4,481 tokens.mdActivating warehouse-modeled data into the tools where work happens. Reverse ETLcited by Reverse ETL and Operational Analytics
- Semantic Layer and Headless BI21 KB · ~5,320 tokens.mdA semantic layer is a centralized, version-controlled, code-defined layer that maps physical warehouse tables to business concepts — entities, dimensions, and metrics — so that 'revenue,' 'active usercited by Semantic Layer and Headless BI
- Survival Analysis15 KB · ~3,965 tokens.mdModeling the time until an event happens when some observations are incomplete (censored or truncated). This is its own discipline because ordinary regression cannot use a row that says 'this customercited by Survival Analysis
- Synthetic Data Generation13 KB · ~3,352 tokens.mdSynthetic data is artificial data produced by a model fit to real data, designed to reproduce the real data's statistical properties (marginals, correlations, joint structure) without being a copy ofcited by Synthetic Data Generation
- Text Analytics and NLP for Analysts16 KB · ~4,198 tokens.mdApplied NLP for turning unstructured text into measurable signal. The audience is an analyst, not an ML engineer: the goal is defensible insight from reviews, support tickets, survey open-ends, call tcited by Text Analytics and NLP for Analysts
A/B Testing Fundamentals29 facts · 5 facets
anomaly detection89 facts · 32 facets
Augmented Analytics and LLM-Assisted Analysis93 facts · 18 facets
Bayesian Data Analysis and Probabilistic Programming70 facts · 11 facets
Causal Discovery and Structure Learning104 facts · 18 facets
Cohort and Retention Analytics132 facts · 22 facets
Conformal Prediction and Uncertainty Quantification80 facts · 23 facets
Customer Lifetime Value Modeling69 facts · 21 facets
Customer-Facing and Embedded Analytics Dashboards70 facts · 21 facets
Data Acquisition and Sampling83 facts · 33 facets
Data Analysis Foundations and Theory53 facts · 15 facets
Data Analysis Lifecycle86 facts · 28 facets
Data FinOps and Cost Optimization89 facts · 15 facets
Data Governance Catalogs and Discovery101 facts · 21 facets
Data Observability106 facts · 18 facets
Data Visualization71 facts · 17 facets
Deep Reinforcement Learning Foundations97 facts · 20 facets
Dimensional and Analytics Data Modeling90 facts · 19 facets
Geospatial Analytics117 facts · 24 facets
Knowledge Graphs and Semantic Analytics40 facts · 10 facets
Machine Learning66 facts · 33 facets
Marketing Mix Modeling and Incrementality86 facts · 18 facets
ML Model Monitoring16 facts · 5 facets
Network and Graph Analytics96 facts · 19 facets
Prescriptive Analytics and Optimization88 facts · 21 facets
Pricing and Revenue Analytics115 facts · 20 facets
Product Analytics127 facts · 22 facets
Real-Time OLAP and Analytical Databases101 facts · 20 facets
Recommender Systems and Learning-to-Rank Analytics92 facts · 17 facets
Reporting and Communication34 facts · 7 facets
Reverse ETL and Operational Analytics109 facts · 22 facets
Semantic Layer and Headless BI100 facts · 16 facets
Survival Analysis113 facts · 26 facets
Synthetic Data Generation119 facts · 24 facets
Text Analytics and NLP for Analysts109 facts · 22 facets
Not yet filed under a root11 files · 19 concepts
- Agent State Management & Durable Execution for LLM Agents26 KB · ~6,600 tokens.mdLLM agents need two distinct things from their runtime: a way to manage state (the data the agent reasons over) and a way to survive failure over long horizons (durable execution). LangGraph is the docited by agent-state-management-durable-execution-for-llm-agents
- Agentic & Advanced RAG Patterns (beyond naive RAG, 2024-2026)22 KB · ~5,528 tokens.mdResearch date: 2026-05-31cited by agentic-advanced-rag-patterns-beyond-naive-rag-2024-2026
- AI Coding-Agent Design (2024–2026)24 KB · ~6,046 tokens.mdConcept: The design discipline of code agents — how autonomous and semi-autonomous LLM systems index code, manage context, apply edits, run control loops, design tools, and get evaluated.cited by ai-coding-agent-design-2024-2026
- AI Red-Teaming & Security-Testing Tooling for LLM Apps (2024-2026)34 KB · ~8,672 tokens.mdAI red-teaming is the offensive testing discipline for LLM and generative-AI applications: systematically generating adversarial inputs to find where a model or app fails in ways we don't want (jailbrcited by ai-red-teaming-security-testing-tooling-for-llm-apps-2024-2026
- AI-Native UX & Generative UI Design for LLM Applications28 KB · ~7,099 tokens.mdAI-native UX is consolidating into a recognizable pattern language. Three findings dominate. First, streaming is the foundational UX primitive: users perceive streaming interfaces as ~40% faster thancited by ai-native-ux-generative-ui-design-for-llm-applications
- Context Engineering for LLM Apps & Agents28 KB · ~7,093 tokens.mdContext engineering is the emerging discipline of curating and maintaining the optimal set of tokens (information) supplied to an LLM at inference time — system prompt, instructions, retrieved documencited by context-engineering-for-llm-apps-agents
- Declarative & Programmatic LLM Frameworks ('Prompt-as-Program', 2024-2026)32 KB · ~8,155 tokens.mdA class of frameworks that emerged in 2023-2024 and matured through 2026 reframes prompting as a software-engineering discipline rather than string authoring. The unifying thesis: don't write prompt scited by declarative-programmatic-llm-frameworks-prompt-as-program-2024-2026
- Eval-Driven Development for LLM Applications (2024-2026)19 KB · ~4,882 tokens.mdResearch date: 2026-05-31cited by eval-driven-development-for-llm-applications-2024-2026
- RLHF & RL Training Infrastructure for LLMs (2024–2026)32 KB · ~8,206 tokens.mdPost-training reinforcement learning (RLHF, RLVR, reasoning-RL, agentic-RL) for LLMs runs on a distinct systems stack that is neither the RL algorithm (PPO/GRPO/DPO) nor generic supervised distributedcited by rlhf-rl-training-infrastructure-for-llms-2024-2026
- The Psychology of Trust, Rapport & Psychological Safety23 KB · ~5,858 tokens.mdThree distinct but interlocking literatures govern professional relationships. Interpersonal trust is best modeled as a willingness to be vulnerable based on perceived ability, benevolence, and integrcited by the-psychology-of-trust-rapport-psychological-safety
- Voice & Real-Time Agent Design (2024–2026)35 KB · ~8,983 tokens.md> Scope note: This report covers the application / design layer of voice agents — architecture choices, real-time APIs, turn-taking, latency engineering, orchestration frameworks, components, UX, andcited by voice-real-time-agent-design-2024-2026
$densify and $fill Gap Filling256 facts · 6 facets
$lookup Patterns228 facts · 6 facets
$merge-$out145 facts · 7 facets
24-hour cooling-off before sentence passes44 facts · 6 facets
2dsphere Geospatial Indexes57 facts · 5 facets
Agent State Management & Durable Execution for LLM Agents85 facts · 20 facets
Agentic & Advanced RAG Patterns (beyond naive RAG, 2024-2026)125 facts · 20 facets
AI Coding-Agent Design (2024–2026)125 facts · 21 facets
AI Red-Teaming & Security-Testing Tooling for LLM Apps (2024-2026)125 facts · 22 facets
AI-Native UX & Generative UI Design for LLM Applications111 facts · 19 facets
Context Engineering for LLM Apps & Agents99 facts · 20 facets
Declarative & Programmatic LLM Frameworks ("Prompt-as-Program", 2024-2026)120 facts · 28 facets
Eval-Driven Development for LLM Applications (2024-2026)99 facts · 19 facets
Heart121 facts · 7 facets
MongoDB Ops Manager62 facts · 19 facets
mongodb-developer181 facts · 40 facets
RLHF & RL Training Infrastructure for LLMs (2024–2026)92 facts · 19 facets
The Psychology of Trust, Rapport & Psychological Safety82 facts · 14 facets
Voice & Real-Time Agent Design (2024–2026)138 facts · 20 facets
Finance, Markets & Blockchain13 files · 15 concepts
- Bitcoin protocol and ecosystem1 KB · ~322 tokens.mdBitcoin protocol and ecosystem expert covering base-layer mechanics, network upgrades, and the L2/metaprotocol frontier. Owns: UTXO model, coin selection algorithms; transaction structure (txid vs wtxcited by Bitcoin protocol and ecosystem
- Budgeting and Saving15 KB · ~3,777 tokens.mdGeneral educational information only — not financial advice. Methods, dollarcited by Budgeting and Saving
- Consumer Finance1 KB · ~289 tokens.mdHub for US personal/consumer finance — money management, coverage, and life planning. Educational only, not financial advice, as of 2026. Routes to 9 spokes.cited by Consumer Finance
- Estate Planning and Wills21 KB · ~5,393 tokens.md> Educational legal information only; NOT legal advice. Estate law is state-specific and fact-specific, and NC statutes change. Everything here is as of 2026 and describes North Carolina law for an incited by Estate Planning and Wills
- Ethereum Consensus (Proof of Stake): Protocol-User Technical Reference50 KB · ~12,783 tokens.md> Scope: protocol-USER level. DEFERRED to siblings (named, not derived): formal BFT safety/liveness proofs and the n>3f bound → distributed-systems-and-consensus; BLS/KZG/VRF crypto internals → cryptocited by Ethereum Consensus (Proof of Stake): Protocol-User Technical Reference
- Health Insurance and Coverage21 KB · ~5,259 tokens.md> Framing — read first. This is **general educational information, NOTcited by Health Insurance and Coverage
- Investing and Retirement19 KB · ~4,778 tokens.md> Framing (read first). This is general educational information, NOT investment, tax, or financial advice, and NOT a recommendation to buy or sell any security. All investing carries risk, including lcited by Investing and Retirement
- Medical Debt and Billing30 KB · ~7,628 tokens.md> Spoke of the consumer-finance hub. This skill covers the practicalcited by Medical Debt and Billing
- Personal Banking22 KB · ~5,575 tokens.md> Framing. This is general educational information, not financial advice.cited by Personal Banking
- Personal Income Taxes24 KB · ~6,102 tokens.md> Educational information, NOT tax advice. This is a practical filer'scited by Personal Income Taxes
- Personal Insurance26 KB · ~6,727 tokens.md> Spoke of the consumer-finance hub. Covers personal lines property/casualtycited by Personal Insurance
- Student Loans15 KB · ~3,874 tokens.md> Framing — read first. This is general educational information, not financial, tax, or legal advice. US student-loan policy is exceptionally volatile (2024-2026): the SAVE plan was struck down, the Ocited by Student Loans
- Trading and Investing — Active Trading & How Financial Markets Work (Family Root)39 KB · ~9,883 tokens.mdThe front door for active trading and how financial markets actually work for a US retail participant. This skill does two jobs:cited by Trading and Investing — Active Trading & How Financial Markets Work (Family Root)
401(k)/403(b) (match/vesting/Roth vs traditional)68 facts · 6 facets
ACA premium tax credits/APTC & CSRs53 facts · 6 facets
Bitcoin protocol and ecosystem17 facts · 5 facets
Budgeting and Saving78 facts · 17 facets
Consumer Finance19 facts · 5 facets
Estate Planning and Wills91 facts · 19 facets
Ethereum Consensus (Proof of Stake): Protocol-User Technical Reference185 facts · 27 facets
Health Insurance and Coverage107 facts · 18 facets
Investing and Retirement105 facts · 17 facets
Medical Debt and Billing154 facts · 33 facets
Personal Banking114 facts · 24 facets
Personal Income Taxes113 facts · 30 facets
Personal Insurance134 facts · 33 facets
Student Loans93 facts · 18 facets
Trading and Investing — Active Trading & How Financial Markets Work (Family Root)45 facts · 14 facets
Sciences & Human Behavior13 files · 13 concepts
- Applied Human Psychology4 KB · ~1,003 tokens.mdTen evidence-based applied-psychology skills, folded as on-demand references. Keep always-loaded index small. Each reference = full former standalone skill; several carry own references/ depth.cited by Applied Human Psychology
- Behavioral Decision-Making and Cognitive Biases15 KB · ~3,865 tokens.mdThe descriptive account of judgment and decision-making: how people actually decide, not how they should. Use it to read why a customer, buyer, or stakeholder made a seemingly irrational choice, to etcited by Behavioral Decision-Making and Cognitive Biases
- Emotion & Affect Psychology36 KB · ~9,271 tokens.mdThe evidence-based science of emotion: what it is, how it arises, how itcited by Emotion & Affect Psychology
- Health Behavior Change and Donor Registration2 KB · ~425 tokens.mdApplied health-behavior change models and organ-donor registration intervention evidence. Covers the major theoretical frameworks used to design health-behavior campaigns, applied specifically to donocited by Health Behavior Change and Donor Registration
- Learning & Expertise Psychology38 KB · ~9,668 tokens.mdThe science of how durable skill and knowledge are actually built — and the uncomfortable headline that organizes the whole field: the activities that feel like learning are usually not, and the activcited by Learning & Expertise Psychology
- Moral Psychology13 KB · ~3,312 tokens.mdThe descriptive science of how people judge right, wrong, and fair — and whycited by Moral Psychology
- Performance, Motivation & Resilience Psychology42 KB · ~10,861 tokens.mdThe evidence-based psychology of sustaining performance over the long haul —cited by Performance, Motivation & Resilience Psychology
- Personality and Individual Differences25 KB · ~6,295 tokens.mdThe evidence-based science of stable individual differences, meaning who acited by Personality and Individual Differences
- Persuasion & Influence Psychology (Attitude-Change Theory)14 KB · ~3,605 tokens.mdThe mechanism layer of attitude change. This skill answers why a persuasion attempt succeeds, fails, or backfires — the cognitive and motivational machinery underneath the tactics. It does not teach ycited by Persuasion & Influence Psychology (Attitude-Change Theory)
- Psychology of Charitable Giving2 KB · ~424 tokens.mdThe empirical psychology of charitable giving — why people give (or don't). Distinct from fundraising-and-donor-psychology (which covers what to DO given they might give) and from psychology-of-charitcited by Psychology of Charitable Giving
- Psychology of Human-AI Interaction (Trust & Appropriate Reliance)10 KB · ~2,636 tokens.md> Standalone skill authored via the /dr deep-research workflow. Full SKILL.mdcited by Psychology of Human-AI Interaction (Trust & Appropriate Reliance)
- Scientific Phylogenetics (ETE Toolkit)6 KB · ~1,628 tokens.mdETE (Environment for Tree Exploration) is a Python toolkit for phylogenetic and hierarchical tree analysis. Core use cases: tree I/O and manipulation, evolutionary event detection, NCBI taxonomy integcited by Scientific Phylogenetics (ETE Toolkit)
- Trust, Rapport & Psychological Safety9 KB · ~2,214 tokens.mdThree distinct but interlocking literatures govern professional relationships. Trust is dyadic willingness to be vulnerable. Rapport is the moment-to-moment relational tone. Psychological safety is acited by Trust, Rapport & Psychological Safety
Applied Human Psychology20 facts · 6 facets
Behavioral Decision-Making and Cognitive Biases70 facts · 17 facets
Emotion & Affect Psychology154 facts · 19 facets
Health Behavior Change and Donor Registration20 facts · 5 facets
Learning & Expertise Psychology136 facts · 20 facets
Moral Psychology51 facts · 17 facets
Performance, Motivation & Resilience Psychology188 facts · 20 facets
Personality and Individual Differences99 facts · 22 facets
Persuasion & Influence Psychology (Attitude-Change Theory)75 facts · 18 facets
Psychology of Charitable Giving25 facts · 5 facets
Psychology of Human-AI Interaction (Trust & Appropriate Reliance)70 facts · 19 facets
Scientific Phylogenetics (ETE Toolkit)44 facts · 14 facets
Trust, Rapport & Psychological Safety25 facts · 9 facets
DevOps, Infrastructure & Observability3 files · 13 concepts
- Linux Boot & Init — UEFI/Secure Boot, GRUB, initramfs/dracut, Early Userspace18 KB · ~4,688 tokens.mdOn a modern machine the boot is a chain of trust and handoffs, each stage finding, optionallycited by Linux Boot & Init — UEFI/Secure Boot, GRUB, initramfs/dracut, Early Userspace
- Linux Package Management & Software Building — apt/dpkg, dnf/rpm, pacman, from-source & kernel build18 KB · ~4,654 tokens.mdA Linux package is an archive of files plus metadata (name, version, dependencies, scripts,cited by Linux Package Management & Software Building — apt/dpkg, dnf/rpm, pacman, from-source & kernel build
- systemd (init system & service manager)1 KB · ~249 tokens.mdDevOps / infrastructure / observability family ROUTER. Split into focused sub-hubs — route to: devops-linux-internals (kernel, boot, memory/NUMA, storage/filesystems, virtualization, io_uring, cgroupscited by DevOps, Infrastructure & Observability, eBPF for Linux Observability, Networking & Security, Immutable & Atomic Linux Distributions — OSTree/rpm-ostree, bootc/CoreOS, NixOS, openSUSE MicroOS +8
DevOps, Infrastructure & Observability2 facts · 1 facets
eBPF for Linux Observability, Networking & Security2 facts · 1 facets
Immutable & Atomic Linux Distributions — OSTree/rpm-ostree, bootc/CoreOS, NixOS, openSUSE MicroOS2 facts · 1 facets
Linux Boot & Init — UEFI/Secure Boot, GRUB, initramfs/dracut, Early Userspace88 facts · 15 facets
Linux Filesystems & Storage — ext4/XFS/Btrfs/ZFS, LVM, Block Layer & I/O Schedulers2 facts · 1 facets
Linux io_uring — Async I/O Rings, liburing, Registered Resources & Security-Disable Saga2 facts · 1 facets
Linux Kernel Architecture & Scheduling — CFS/EEVDF, Syscall ABI, Kernel Modules2 facts · 1 facets
Linux Mandatory Access Control & Privilege — SELinux, AppArmor & Capabilities2 facts · 1 facets
Linux Memory Management & NUMA — Virtual Memory, Paging, Reclaim, OOM, Hugepages, NUMA Tuning2 facts · 1 facets
Linux Package Management & Software Building — apt/dpkg, dnf/rpm, pacman, from-source & kernel build84 facts · 16 facets
Linux Sandboxing & Confinement — seccomp-bpf, Landlock, gVisor, Kata, Firecracker2 facts · 1 facets
Linux Virtualization — KVM, QEMU, libvirt, virtio & microVMs2 facts · 1 facets
systemd (init system & service manager)2 facts · 1 facets
Personal Life & Property0 files · 11 concepts
Auction Vehicle Repair Cost Estimation156 facts · 8 facets
AutoBidMaster International Bidding194 facts · 6 facets
Best-Value Used Car Selection116 facts · 6 facets
Copart Salvage Auctions225 facts · 6 facets
IAAI Salvage Auctions150 facts · 8 facets
Individual Buyer Best Practices at Salvage Auctions124 facts · 6 facets
Low-Cost Alternative Car-Buying Channels47 facts · 7 facets
Newbie Mistakes in Used Car Buying90 facts · 7 facets
Reading a Salvage Auction Listing172 facts · 7 facets
Salvage and Clean-Title Vehicle Valuation264 facts · 8 facets
Salvage Auction Bidding Strategy126 facts · 7 facets
Business & Enterprise5 files · 5 concepts
- Personal Venture — NC Organ-Donation Nonprofit & Founder Toolkit33 KB · ~8,399 tokens.mdEducational domain reference for a North Carolina cause venture working on organ, eye, and tissue donation awareness and donor registration. It explains how the system is structured and governed, howcited by Personal Venture — NC Organ-Donation Nonprofit & Founder Toolkit
- Semantic Monitoring, Reporting & Customer Dashboards for TAM (value-chain synthesis)4 KB · ~924 tokens.mdHub for the MongoDB Technical Account Manager's operational toolkit — producing account deliverables, scoring customer health, running case and incident operations, automating reports, and integratingcited by Semantic Monitoring, Reporting & Customer Dashboards for TAM (value-chain synthesis)
- TAM commercial metrics12 KB · ~3,104 tokens.mdThis reference is the numbers-and-definitions layer for a TAM. It answers three recurring questions:cited by TAM commercial metrics
- TAM Expertise2 KB · ~595 tokens.mdReference skill compiled from 120+ authoritative sources. Full context in tam-expertise-context.md.cited by TAM Expertise
- Value Realization and Outcome-Based Customer Success13 KB · ~3,247 tokens.md> Reference skill — part of the tam-operations family (the 'Semantic Monitoring → Reporting → Dashboards → TAM Methodology' value-chain synthesis). This file owns the practice of defining, driving, mecited by Value Realization and Outcome-Based Customer Success
Personal Venture — NC Organ-Donation Nonprofit & Founder Toolkit143 facts · 21 facets
Semantic Monitoring, Reporting & Customer Dashboards for TAM (value-chain synthesis)32 facts · 10 facets
TAM commercial metrics73 facts · 21 facets
TAM Expertise52 facts · 13 facets
Value Realization and Outcome-Based Customer Success74 facts · 21 facets