research-lab
Skill4 starsChanged 6 months agoMITView on GitHub
Full-stack AI research and data science AI covering 20+ specialist skills across data science, machine learning, AI agent architecture, multi-agent orchestration, LLM engineering, RAG systems, voice AI, prompt engineering, and scientific research methodology. Use for data analysis, predictive modeling, statistical inference, A/B testing, autonomous agent design, ReAct loops, multi-agent coordination (CrewAI, LangGraph), agent memory systems, RAG pipelines, prompt optimization, voice agents, agent evaluation, and research engineering. Trigger keywords: data science, machine learning, ML model, statistical analysis, A/B test, predictive model, AI agent, autonomous agent, multi-agent, orchestration, CrewAI, LangGraph, RAG, retrieval-augmented generation, prompt engineering, LLM app, agent memory, voice agent, voice AI, research engineer, scientific research, data scientist, computer use agent, agent evaluation, brainstorming, ideation.
3 other projects use a version of this.
Adopted by
Most recently taken up by
- ooooooooooooooooooop/agent-toolstheir own version, first seen Oct 22, 2026
- ooooooooooooooooooop/personal-aitheir own version, first seen Oct 22, 2026
- Xplore-LAB/openclaw-scholar-packtheir own version, first seen Apr 2, 2026
Other versions3 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 3
- ooooooooooooooooooop/agent-tools★ 4same name
- ooooooooooooooooooop/personal-ai★ 4same name
- Xplore-LAB/openclaw-scholar-pack★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository10 other files it gives its agents.
- all-marketingSkill
- coding-suiteSkill
- design-studioSkill
- devops-suiteSkill
- dispatchSkill
- founders-suiteSkill
- game-devSkill
- marketing-suiteSkill
- productivity-suiteSkill
- purple-teamSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 80
All sub-skill instructions are in references/skills-catalog.md.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Research Lab
A complete, modular AI research and engineering suite — from rigorous data science and statistical modeling to cutting-edge AI agent architecture, multi-agent systems, LLM engineering, and voice AI.
How to Use This Skill
- Identify the task category from the routing table below
- Load the matching sub-skill from
references/skills-catalog.md - Execute following the sub-skill's workflow
Quick Routing Table
📊 Data Science & Research
| Task | Load |
| Data analysis, ML modeling, statistical inference, A/B testing, visualization | data-scientist |
| Rigorous academic research engineering — formal correctness, HPC, proofs | research-engineer |
| Scientific methodology, research design, literature synthesis | scientific-research |
| Brainstorming, ideation, structured creative thinking | brainstorming |
🤖 AI Agent Architecture
| Task | Load |
| Designing autonomous AI agents — tool use, memory, planning, guardrails | ai-agents-architect |
| Autonomous agent loops — ReAct, Plan-Execute, reflection, self-correction | autonomous-agents |
| Reusable agent design patterns — subagent coordination, handoffs | autonomous-agent-patterns |
| Computer use agents — GUI automation, screen interaction, browser agents | computer-use-agents |
🌐 Multi-Agent Systems
| Task | Load |
| Multi-agent orchestration — optimizing performance across agent networks | agent-orchestration |
| Dispatching and routing tasks to parallel agents | dispatching-parallel-agents |
| Subagent-driven development — building with specialized subagents | subagent-driven-dev |
| CrewAI — role-based crew design, task delegation, agent collaboration | crewai |
| LangGraph — stateful agent graphs, conditional edges, human-in-the-loop | langgraph |
🧠 Memory & State
| Task | Load |
| Agent memory systems — episodic, semantic, procedural, long-term persistence | agent-memory-systems |
| MCP-based memory integration for agents | agent-memory-mcp |
⚗️ LLM Engineering
| Task | Load |
| Prompt engineering patterns — few-shot, CoT, structured output, debugging | prompt-engineering |
| Prompt library management — versioning, evaluation, organization | prompt-library |
| Prompt caching strategies — cost reduction, latency optimization | prompt-caching |
| RAG pipelines — chunking, embedding, retrieval, reranking, evaluation | rag-engineer |
| LLM application patterns — streaming, tool use, error handling, evals | llm-app-patterns |
🎙️ Voice & Multimodal AI
| Task | Load |
| Voice agent design — STT/TTS pipelines, latency, interruption handling | voice-agents |
| Voice AI development — real-time audio, WebRTC, telephony integration | voice-ai-development |
🔬 Evaluation
| Task | Load |
| Agent evaluation — benchmarks, metrics, failure analysis, test harnesses | agent-evaluation |
Loading Sub-Skills
All sub-skill instructions are in references/skills-catalog.md.
research-lab/
├── SKILL.md ← You are here (routing hub)
└── references/
└── skills-catalog.md ← Full instructions for all skillsAlways read the relevant section of skills-catalog.md before executing any task.
Skill Sorting Notes
Skills from this folder that route to other suites:
analytics-tracking→ founders-suite (Segment CDP / event tracking)architecture-decision-records→ devops-suitebullmq-specialist→ devops-suite (message queue infrastructure)email-sequence→ marketing-suiteemail-systems→ coding-suiteemployment-contract-templates→ founders-suite (HR workflows)on-call-handoff-patterns→ devops-suiteplaid-fintech→ coding-suite (API integration)kaizen→ founders-suite (process improvement framework)
Universal Research Lab Standards
Correctness before performance — A fast wrong answer is worse than a slow right one. Validate assumptions, cite sources, acknowledge uncertainty.
Reproducibility is non-negotiable — Every experiment, prompt, and pipeline must be reproducible. Version your prompts, seed your random generators, log your hyperparameters.
Agents earn autonomy — Start constrained. Add autonomy only when reliability is proven. A 95% step success rate = 60% success by step 10.
Evaluate everything — Build evals before features. If you can't measure it, you can't improve it. Every LLM pipeline needs a test harness.
Data integrity first — Garbage in, garbage out. Understand your data distribution, validate your splits, and document your preprocessing.
Complexity when necessary, not as default — The simplest model that solves the problem is the right model. Add complexity only when it demonstrably improves outcomes.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.