ideate

Academic Writer — Research Ideation

Guided ideation flow that helps researchers formulate strong research questions before writing.

Load Profile

cat .academic-helper/profile.md

If no profile, tell the researcher to run /academic-writer:init first.

Extract fieldOfStudy, targetLanguage, and tools configuration.

Cognetivy Tracking

If Cognetivy is enabled, start a run:

echo '{"phase": "ideation"}' > /tmp/aw-ideate-input.json
cognetivy run start --workflow wf_ideate --input /tmp/aw-ideate-input.json

Record the run ID for all logging below.

---

Step 1: Topic Exploration

Ask the researcher:

AskUserQuestion(questions=[{
  "question": "What topic or area are you thinking about exploring? This can be vague — we'll sharpen it together.",
  "header": "Step 1 — Topic Exploration",
  "options": []
}])

Show examples relevant to their fieldOfStudy:

"Examples:

• A specific text or passage you've been thinking about
• A debate in your field you want to contribute to
• A concept that seems under-explored
• Something you noticed in your sources that doesn't fit existing explanations"

---

Step 2: 5W1H Brainstorming

Apply the 5W1H framework to the topic. Generate questions in targetLanguage:

Dimension	Question	Purpose
What (מה)	What is the phenomenon, text, or concept?	Define the object of study
Who (מי)	Who are the key figures, authors, or communities?	Identify actors
When (מתי)	What is the historical period or textual dating?	Temporal context
Where (איפה)	What is the geographic, cultural, or textual context?	Spatial/contextual frame
Why (למה)	Why does this matter? Why now?	Significance and urgency
How (כיצד)	How does this work? How was it received/transmitted?	Mechanism or process

Present the brainstorm:

"Here's a 5W1H exploration of your topic:

What: [...] Who: [...] When: [...] Where: [...] Why: [...] How: [...]

Which angles interest you most?"

Log to Cognetivy:

echo '{"type":"step_completed","nodeId":"brainstorm_5w1h"}' | cognetivy event append --run RUN_ID

---

Step 3: Gap Analysis

If Candlekeep is enabled, scan the researcher's library:

ck items list --json

For each source, check its metadata against the topic:

• What has been written about this topic in the researcher's sources?
• What perspectives or arguments are represented?
• What's missing — which angles are NOT covered?

If Agentic-Search-Vectorless is enabled, query for the topic:

bash plugins/academic-writer/scripts/vectorless-query.sh --query "TOPIC_DESCRIPTION" --mode global --top-k 30

Present the gap analysis:

"Based on your source library:

Covered ground:

• [Source A] addresses [aspect X]
• [Source B] discusses [aspect Y]

Gaps I identified:

• No source addresses [aspect Z]
• The relationship between [X] and [Y] is unexplored
• [Historical period / figure / concept] is mentioned but never analyzed

These gaps could be fertile ground for a research question."

Log:

echo '{"type":"step_completed","nodeId":"gap_analysis","sourcesScanned":N,"gapsFound":N}' | cognetivy event append --run RUN_ID

---

Step 4: Approach Framing

Present Humanities-appropriate research approaches:

AskUserQuestion(questions=[{
  "question": "What kind of study fits your topic best?",
  "header": "Step 4 — Research Approach",
  "options": [
    {
      "label": "Textual Analysis",
      "description": "Close reading of a specific text or corpus",
      "markdown": "```\nTextual Analysis\n────────────────\nFocus: One or more primary texts\nMethod: Close reading, literary/philological analysis\nOutput: New interpretation or reading\nBest for: Discovering meaning in texts\n```"
    },
    {
      "label": "Comparative Study",
      "description": "Comparing two or more texts, figures, or traditions",
      "markdown": "```\nComparative Study\n─────────────────\nFocus: Two+ texts, thinkers, or traditions\nMethod: Systematic comparison on defined axes\nOutput: Reveals connections or tensions\nBest for: Cross-cultural or cross-period analysis\n```"
    },
    {
      "label": "Historical Survey",
      "description": "Tracing the development of an idea, practice, or text over time",
      "markdown": "```\nHistorical Survey\n─────────────────\nFocus: An idea or practice across time\nMethod: Chronological analysis with sources\nOutput: Developmental narrative\nBest for: Showing change over time\n```"
    },
    {
      "label": "Conceptual Analysis",
      "description": "Analyzing a philosophical, theological, or cultural concept",
      "markdown": "```\nConceptual Analysis\n───────────────────\nFocus: A concept or term\nMethod: Definition, usage mapping, critical analysis\nOutput: Clarified or redefined concept\nBest for: Philosophical and theological work\n```"
    },
    {
      "label": "Reception History",
      "description": "How a text or idea was received, interpreted, and transformed",
      "markdown": "```\nReception History\n─────────────────\nFocus: How a text was read across generations\nMethod: Track interpretive traditions\nOutput: Map of readings and their contexts\nBest for: Exegetical and hermeneutical studies\n```"
    },
    {
      "label": "Other",
      "description": "Describe your own approach",
      "markdown": "```\nCustom Approach\n───────────────\n→ Describe your methodology\n→ I'll help frame it\n```"
    }
  ],
  "multiSelect": false
}])

---

Step 5: Research Question Formulation

Based on the topic, gaps, and approach, propose 3 structured research questions in targetLanguage:

For each question, provide:

1. The question — clear, specific, answerable
2. Thesis direction — what the likely argument would be
3. Key sources — which of the researcher's sources are most relevant
4. Feasibility — can this be answered with available sources?

Present and ask:

AskUserQuestion(questions=[{
  "question": "Here are three research questions I've formulated. Which one excites you?",
  "header": "Research Questions",
  "options": [
    {"label": "Question 1", "description": "SHORT_SUMMARY_1"},
    {"label": "Question 2", "description": "SHORT_SUMMARY_2"},
    {"label": "Question 3", "description": "SHORT_SUMMARY_3"},
    {"label": "I want to refine one", "description": "Tell me which to adjust"},
    {"label": "None — let me describe my own", "description": "Free-text input"}
  ],
  "multiSelect": false
}])

---

Step 6: Generate Research Brief

Write a research-brief.md file that feeds directly into /academic-writer:write:

mkdir -p .academic-helper

Use the Write tool to create .academic-helper/research-brief.md:

# Research Brief

## Research Question
[The approved question]

## Thesis Direction
[Preliminary thesis statement]

## Approach
[Selected methodology]

## Key Sources
1. [Source with ID] — relevance
2. [Source with ID] — relevance
3. ...

## Gap This Addresses
[What's missing in existing scholarship]

## Scope
- Estimated sections: [N]
- Primary texts: [list]
- Secondary literature: [list]

## 5W1H Summary
- What: [...]
- Who: [...]
- When: [...]
- Where: [...]
- Why: [...]
- How: [...]

---
*Generated by /academic-writer:ideate on [DATE]*
*Feed this into /academic-writer:write to start writing.*

Log completion:

echo '{"type":"step_completed","nodeId":"research_brief","outputPath":".academic-helper/research-brief.md"}' | cognetivy event append --run RUN_ID
cognetivy run complete --run RUN_ID

Completion

"Your research brief has been saved to .academic-helper/research-brief.md.

Research question: [question] Approach: [approach] Key sources: [N] identified

To start writing, run /academic-writer:write. The brief will be loaded automatically as a starting point."