home / skills / simhacker / moollm / research-notebook

research-notebook skill

safe

This skill helps you organize structured research notebooks with questions, sources, findings, and decisions to improve knowledge capture.

npx playbooks add skill simhacker/moollm --skill research-notebook

Review the files below or copy the command above to add this skill to your agents.

Files (7)

SKILL.md

3.0 KB

---
name: research-notebook
description: "Structured research with sources, findings, and decisions."
license: MIT
tier: 1
allowed-tools:
  - read_file
  - write_file
related: [play-learn-lift, summarize, memory-palace, debugging, scratchpad, session-log, sister-script]
tags: [moollm, research, investigation, sources, findings]
---

# Research Notebook

> **Structured research with sources, findings, and decisions.**

Track investigations and capture knowledge.

> [!TIP]
> **LEARN stage of [play-learn-lift](../play-learn-lift/).** Where patterns emerge from exploration.

## Structure

| Section | Purpose |
|---------|---------|
| **Questions** | What you're trying to learn |
| **Sources** | Where you looked |
| **Findings** | What you discovered |
| **Decisions** | What you concluded |

## Contents

| File | Purpose |
|------|---------|
| [SKILL.md](./SKILL.md) | Full protocol documentation |
| [NOTEBOOK.yml.tmpl](./NOTEBOOK.yml.tmpl) | Notebook template |
| [SOURCES.yml.tmpl](./SOURCES.yml.tmpl) | Sources template |

## Example

```yaml
research:
  topic: "LLM context management strategies"
  
  questions:
    - "How do other systems handle token budgets?"
    - "What summarization approaches exist?"
    
  sources:
    - url: "paper-on-context.pdf"
      relevance: "Direct answer to Q1"
      
  findings:
    - "Sliding window is most common"
    - "Hierarchical summaries preserve more"
    
  decisions:
    - "Use tiered hot/cold approach"
```

## The Intertwingularity

Research-notebook is the LEARN stage of [play-learn-lift](../play-learn-lift/) — find patterns.

```mermaid
graph LR
    RN[📓 research-notebook] -->|LEARN stage of| PLL[🎮📚🚀 play-learn-lift]
    RN -->|distills to| SUM[📝 summarize]
    RN -->|sources from| SL[📜 session-log]
    RN -->|informs| SS[👯 sister-script]
    
    DB[🔧 debugging] -->|uses| RN
    AP[⚔️ adventure] -->|evidence in| RN
```

---

## Dovetails With

### Sister Skills
| Skill | Relationship |
|-------|--------------|
| [play-learn-lift/](../play-learn-lift/) | Research IS the LEARN stage |
| [session-log/](../session-log/) | Source material from PLAY |
| [summarize/](../summarize/) | Compress findings |
| [sister-script/](../sister-script/) | Research feeds automation |
| [debugging/](../debugging/) | Research supports hypothesis tracking |
| [adventure/](../adventure/) | Evidence collection |
| [adversarial-committee/](../adversarial-committee/) | **Multiple perspectives** on research questions |
| [rubric/](../rubric/) | Criteria for evaluating sources and findings |

### Protocol Symbols
| Symbol | Link |
|--------|------|
| `PLAY-LEARN-LIFT` | [PROTOCOLS.yml](../../PROTOCOLS.yml#PLAY-LEARN-LIFT) |
| `PRESERVE-JOURNEY` | [PROTOCOLS.yml](../../PROTOCOLS.yml#PRESERVE-JOURNEY) |

### Navigation
| Direction | Destination |
|-----------|-------------|
| ⬆️ Up | [skills/](../) |
| ⬆️⬆️ Root | [Project Root](../../) |
| 🎮 Sister | [play-learn-lift/](../play-learn-lift/) |
| 📝 Sister | [summarize/](../summarize/) |

Overview

This skill organizes investigative work into a concise research notebook that captures questions, sources, findings, and decisions. It provides a lightweight template and structure to track evidence, surface patterns, and make decisions during exploratory work. The focus is on turning scattered observations into actionable conclusions.

How this skill works

You record the research topic and list the core questions driving the investigation. For each question you log sources you consulted with relevance notes, extract concrete findings, and record decisions or next steps informed by evidence. Templates for the notebook and sources standardize entries so research is consistent and reviewable.

When to use it

During exploratory research to collect and organize evidence
When you need to preserve the reasoning behind decisions
While evaluating multiple approaches or alternatives
To capture learnings from experiments, debugging sessions, or user studies
When consolidating material before summarization or handoff

Best practices

State 2–5 clear research questions up front to focus work
Annotate each source with relevance and a short rationale
Capture findings as concise, evidence-backed statements (what, why, confidence)
Record decisions immediately with linked findings or sources to preserve traceability
Keep entries short and frequent — prefer many small updates over long, infrequent reports

Example use cases

Investigating LLM context-management patterns and summarization strategies
Tracking sources and conclusions during debugging or root-cause analysis
Collecting evidence from user interviews to inform product decisions
Evaluating trade-offs between implementation alternatives before committing
Feeding distilled findings into summaries, automations, or design documents

FAQ

How detailed should sources entries be?

Include enough metadata to find the source later (title, URL, file name), a one-line relevance note, and any immediate quotes or timestamps you relied on.

How do findings differ from decisions?

Findings are observations or conclusions supported by evidence. Decisions are choices you make based on those findings, often including next steps and confidence levels.