# Research by task Common AI and LLM tasks, each paired with the source-linked facts files to reference in a prompt, the articles and how-to recipes that show the work, and the skills that do it. Data curated 2026-10-30; twin built 2026-11-02. What this section is and how it is built: [Concept facts files](https://llms-explorer.com/reference/context-files/). ## What this section holds (82) - [Write better prompts and reusable skills · Prompt Helper and Optimizer — facts](https://llms-explorer.com/downloads/concepts/prompt-helper-and-optimizer.md) - [Write better prompts and reusable skills · Instruction-file hygiene for LLM lexical tics: linting, mirroring, hook filtering — facts](https://llms-explorer.com/downloads/concepts/ai-isms-instruction-hygiene.md) - [Write better prompts and reusable skills · Claude Code Skills — facts](https://llms-explorer.com/downloads/concepts/claude-code-skills.md) - [Write better prompts and reusable skills · Claude Code Plugins — facts](https://llms-explorer.com/downloads/concepts/claude-code-plugins.md) - [Write better prompts and reusable skills · Technical paper: Semantic Skill Discovery and the Optimizer Family](https://llms-explorer.com/blog/semantic-skill-discovery-and-the-optimizer-family/) - [Write better prompts and reusable skills · skill: prompt-helper-optimizer](https://llms-explorer.com/skills/prompt-helper-optimizer/) - [Write better prompts and reusable skills · skill: prompt-deep-optimizer](https://llms-explorer.com/skills/prompt-deep-optimizer/) - [Write better prompts and reusable skills · skill: skill-optimizer](https://llms-explorer.com/skills/skill-optimizer/) - [Write better prompts and reusable skills · skill: artifact-to-procedure](https://llms-explorer.com/skills/artifact-to-procedure/) - [Give an agent the right context and memory · LLM Context Engineering — facts](https://llms-explorer.com/downloads/concepts/llm-context-engineering.md) - [Give an agent the right context and memory · Context-Window Pressure & Compaction — facts](https://llms-explorer.com/downloads/concepts/context-window-pressure-compaction.md) - [Give an agent the right context and memory · Agent State & Durable Execution — facts](https://llms-explorer.com/downloads/concepts/agent-state-durable-execution.md) - [Give an agent the right context and memory · Architecture report: State That Survives the Session](https://llms-explorer.com/blog/state-that-survives-the-session/) - [Give an agent the right context and memory · Technical review: Building Codebases for Machine Collaborators](https://llms-explorer.com/blog/building-codebases-for-machine-collaborators/) - [Give an agent the right context and memory · skill: memory-to-llms-txt](https://llms-explorer.com/skills/memory-to-llms-txt/) - [Give an agent the right context and memory · skill: notes-to-llms-txt](https://llms-explorer.com/skills/notes-to-llms-txt/) - [Give an agent the right context and memory · skill: braindump](https://llms-explorer.com/skills/braindump/) - [Cut token cost and latency · Prompt caching — facts](https://llms-explorer.com/downloads/concepts/prompt-caching.md) - [Cut token cost and latency · LLM Inference Optimization and Serving — facts](https://llms-explorer.com/downloads/concepts/llm-inference-optimization-and-serving.md) - [Cut token cost and latency · LLM Model Routing, Cascades & Mixture-of-Agents — facts](https://llms-explorer.com/downloads/concepts/llm-model-routing-cascades-mixture-of-agents.md) - [Cut token cost and latency · Agent skill-listing injection cost — facts](https://llms-explorer.com/downloads/concepts/agent-skill-listing-injection-cost.md) - [Cut token cost and latency · Sourced inventory: Every Token-Saving Strategy in My Stack, With Sources and Numbers](https://llms-explorer.com/blog/every-token-saving-strategy/) - [Cut token cost and latency · Controlled experiment: What Good Docs, llms Files and Indexes Actually Save: 256 Agent Runs](https://llms-explorer.com/blog/what-docs-llms-files-and-indexes-actually-save/) - [Cut token cost and latency · Decision table: Which layer answers which question](https://llms-explorer.com/examples/decision-table/) - [Ground answers in your own documents · Agentic and Advanced RAG Patterns — facts](https://llms-explorer.com/downloads/concepts/agentic-and-advanced-rag-patterns.md) - [Ground answers in your own documents · Hybrid Score Fusion (Reciprocal Rank Fusion - RRF) — facts](https://llms-explorer.com/downloads/concepts/hybrid-score-fusion-reciprocal-rank-fusion-rrf.md) - [Ground answers in your own documents · Cross-Encoder Re-ranking vs Bi-Encoders for Terminal Validation in RAG and Deduplication — facts](https://llms-explorer.com/downloads/concepts/cross-encoder-re-ranking-vs-bi-encoders-for-terminal-validation-in-rag-and-deduplication.md) - [Ground answers in your own documents · LlamaParse vs. Docling for LLM Ingestion — facts](https://llms-explorer.com/downloads/concepts/llamaparse-vs-docling-for-llm-ingestion.md) - [Ground answers in your own documents · Implementation report: Keyword plus vector: the cheap path](https://llms-explorer.com/blog/keyword-plus-vector/) - [Ground answers in your own documents · Recorded walkthrough: Semantic indexing: two legs and a fusion](https://llms-explorer.com/blog/semantic-indexing/) - [Ground answers in your own documents · How-to recipe: Recipe 04 — Hybrid: keyword and vector fused](https://llms-explorer.com/examples/recipe-04/) - [Ground answers in your own documents · How-to recipe: Recipe 07 — Facts into a RAG store](https://llms-explorer.com/examples/recipe-07/) - [Ground answers in your own documents · skill: crawl-to-llms-txt](https://llms-explorer.com/skills/crawl-to-llms-txt/) - [Ground answers in your own documents · skill: llms-concept-abstractor](https://llms-explorer.com/skills/llms-concept-abstractor/) - [Build an agent that plans, acts and recovers · Durable Agent Execution & Long-Running Agent Runtimes — facts](https://llms-explorer.com/downloads/concepts/durable-agent-execution-long-running-agent-runtimes.md) - [Build an agent that plans, acts and recovers · Agent Runtime Sandboxes & Code Execution — facts](https://llms-explorer.com/downloads/concepts/agent-runtime-sandboxes-code-execution.md) - [Build an agent that plans, acts and recovers · Loop & No-Progress Detection — facts](https://llms-explorer.com/downloads/concepts/loop-no-progress-detection.md) - [Build an agent that plans, acts and recovers · AI Coding-Agent Design — facts](https://llms-explorer.com/downloads/concepts/ai-coding-agent-design.md) - [Build an agent that plans, acts and recovers · Annotated transcript: The Deep Document Optimizer — A Live Run](https://llms-explorer.com/blog/the-deep-document-optimizer-a-live-run/) - [Build an agent that plans, acts and recovers · skill: agent-deep-optimizer](https://llms-explorer.com/skills/agent-deep-optimizer/) - [Build an agent that plans, acts and recovers · skill: artifact-to-procedure](https://llms-explorer.com/skills/artifact-to-procedure/) - [Connect an agent to tools with MCP · MCP Server Development — facts](https://llms-explorer.com/downloads/concepts/mcp-server-development.md) - [Connect an agent to tools with MCP · MCP Server Builder Patterns — facts](https://llms-explorer.com/downloads/concepts/mcp-server-builder-patterns.md) - [Connect an agent to tools with MCP · Ollama MCP: MCP clients and servers on local models — facts](https://llms-explorer.com/downloads/concepts/ollama-mcp-mcp-clients-and-servers-on-local-models.md) - [Connect an agent to tools with MCP · MCP tool-definition token cost in OpenCode — facts](https://llms-explorer.com/downloads/concepts/mcp-tool-definition-token-cost-in-opencode.md) - [Connect an agent to tools with MCP · A2A Protocol Interoperability — facts](https://llms-explorer.com/downloads/concepts/a2a-protocol-interoperability.md) - [Connect an agent to tools with MCP · How-to recipe: Recipe 05 — Index-first agent over MCP](https://llms-explorer.com/examples/recipe-05/) - [Connect an agent to tools with MCP · How-to recipe: Recipe 03 — Keyword layer from Claude Code](https://llms-explorer.com/examples/recipe-03/) - [Test and red-team an LLM app · Eval-Driven Development for LLM Apps — facts](https://llms-explorer.com/downloads/concepts/eval-driven-development-for-llm-apps.md) - [Test and red-team an LLM app · LLM-as-judge bias & calibration (kappa, binary-vs-Likert, position/verbosity/self-preference) — facts](https://llms-explorer.com/downloads/concepts/llm-as-judge-bias-calibration-kappa-binary-vs-likert-position-verbosity-self-preference.md) - [Test and red-team an LLM app · AI Red-Teaming and Security-Testing Tooling — facts](https://llms-explorer.com/downloads/concepts/ai-red-teaming-and-security-testing-tooling.md) - [Test and red-team an LLM app · Structured Output Constraints and LLM Hallucination Mitigation — facts](https://llms-explorer.com/downloads/concepts/structured-output-constraints-and-llm-hallucination-mitigation.md) - [Test and red-team an LLM app · Experiment: How many cheap research runs equal one expensive one?](https://llms-explorer.com/blog/how-many-cheap-research-runs-equal-one-expensive-one/) - [Test and red-team an LLM app · skill: deep-eval-optimizer](https://llms-explorer.com/skills/deep-eval-optimizer/) - [Run a model on your own hardware · On-Device & Local LLM Runtimes — facts](https://llms-explorer.com/downloads/concepts/on-device-local-llm-runtimes.md) - [Run a model on your own hardware · Apple silicon unified memory and LLM sizing — facts](https://llms-explorer.com/downloads/concepts/apple-silicon-unified-memory-and-llm-sizing.md) - [Run a model on your own hardware · Tool-call parser and chat-template mismatch — facts](https://llms-explorer.com/downloads/concepts/tool-call-parser-and-chat-template-mismatch.md) - [Run a model on your own hardware · Thunderbolt eGPU on Linux for local LLM inference — facts](https://llms-explorer.com/downloads/concepts/thunderbolt-egpu-linux.md) - [Run a model on your own hardware · Benchmark report: Choosing a local LLM for a 16 GB RTX 5080, part 1: what the bake-off shows so far](https://llms-explorer.com/blog/choosing-a-local-llm-for-a-16gb-rtx-5080-part-1/) - [Run a model on your own hardware · Benchmark report: Choosing a local LLM for a 16 GB RTX 5080, part 2: the winner and how we serve it](https://llms-explorer.com/blog/choosing-a-local-llm-for-a-16gb-rtx-5080-part-2/) - [Run a model on your own hardware · Measured case study: Getting Gemma 4 MLX to do real research on a Mac](https://llms-explorer.com/blog/getting-gemma-mlx-to-do-real-research/) - [Run a model on your own hardware · How-to recipe: Recipe 10 — A local hub in miniature](https://llms-explorer.com/examples/recipe-10/) - [Make your docs readable by agents · llms.txt — facts](https://llms-explorer.com/downloads/concepts/llms-txt.md) - [Make your docs readable by agents · agents.md and the Universal Commerce Protocol (UCP) — facts](https://llms-explorer.com/downloads/concepts/agents-md-ucp.md) - [Make your docs readable by agents · NLWeb and MCP as agentic-discovery alternatives to llms.txt — facts](https://llms-explorer.com/downloads/concepts/nlweb-mcp-agentic-discovery.md) - [Make your docs readable by agents · robots.txt and the Content-Signal AI-preference extension — facts](https://llms-explorer.com/downloads/concepts/robots-txt-content-signals.md) - [Make your docs readable by agents · Recorded survey: Six months of hand-made llms files](https://llms-explorer.com/blog/six-months-of-hand-made-llms/) - [Make your docs readable by agents · How-to recipe: Recipe 01 — Two hops with requests](https://llms-explorer.com/examples/recipe-01/) - [Make your docs readable by agents · How-to recipe: Recipe 09 — Serving with the right headers](https://llms-explorer.com/examples/recipe-09/) - [Make your docs readable by agents · skill: crawl-to-llms-txt](https://llms-explorer.com/skills/crawl-to-llms-txt/) - [Make your docs readable by agents · skill: crawl-repo-to-llms](https://llms-explorer.com/skills/crawl-repo-to-llms/) - [Make your docs readable by agents · skill: notes-to-llms-txt](https://llms-explorer.com/skills/notes-to-llms-txt/) - [Make your docs readable by agents · skill: llms-deep-optimizer](https://llms-explorer.com/skills/llms-deep-optimizer/) - [Research a topic with an agent · Deep Research Methods — facts](https://llms-explorer.com/downloads/concepts/deep-research-methods.md) - [Research a topic with an agent · Deep Research — facts](https://llms-explorer.com/downloads/concepts/deep-research.md) - [Research a topic with an agent · Concept knowledge architecture strategies — facts](https://llms-explorer.com/downloads/concepts/concept-knowledge-architecture-strategies.md) - [Research a topic with an agent · Extraction Evaluation Frameworks — facts](https://llms-explorer.com/downloads/concepts/extraction-evaluation-frameworks.md) - [Research a topic with an agent · Experiment: How many cheap research runs equal one expensive one?](https://llms-explorer.com/blog/how-many-cheap-research-runs-equal-one-expensive-one/) - [Research a topic with an agent · Evaluation report: Abstracting one concept out of many docsets](https://llms-explorer.com/blog/abstracting-one-concept/) - [Research a topic with an agent · skill: /dr — deep-research](https://llms-explorer.com/skills/dr/) - [Research a topic with an agent · skill: concept-family-explorer](https://llms-explorer.com/skills/concept-family-explorer/) - [Research a topic with an agent · skill: rabbithole](https://llms-explorer.com/skills/rabbithole/)