synthesize

name	synthesize
description	Cross-references multiple research documents to identify themes, gaps, conflicts, and priority rankings

Synthesize

Analyze multiple research files together to identify cross-cutting themes, knowledge gaps, conflicts, and priorities.

Inputs

• {{RESEARCH_DIR}} — Directory containing all research markdown files.
• {{OUTPUT_PATH}} — File path where the synthesis report will be written.

Process

1. Discover and read all research files.

   • Use Glob to find all markdown files in {{RESEARCH_DIR}} (pattern: **/*.md).
   • Read each file. Track the topic ID, title, and depth of content for each.

2. Identify cross-cutting themes.

   • Use Grep to search for recurring concepts, terms, and patterns across files.
   • Group topics that share common underlying principles.
   • Note which concepts appear in 3+ different topic areas — these are likely high-priority.

3. Detect knowledge gaps.

   • For each research file, assess whether it has sufficient depth:
     • Does it cover the Overview, Key Concepts, and Technical Details adequately?
     • Are there sections that are thin or vague?
     • Are there referenced subtopics that have no dedicated research file?
   • Flag topics that need additional research.

4. Find conflicting information.

   • Look for cases where different research files make contradictory claims.
   • Note version-specific differences that could cause confusion.
   • Identify areas where "it depends" is the real answer and the nuance needs capturing.

5. Build a dependency graph.

   • Determine which topics build on others (prerequisites).
   • Identify foundational topics that should be learned first.
   • Note topics that are relatively independent and can be approached in any order.

6. Rank priorities.

   • Consider: importance weight, current depth of research, complexity, number of dependencies.
   • Assign each topic area a priority: critical, high, medium, or low.

7. Write the synthesis.

Output Format

Write a markdown file to {{OUTPUT_PATH}} with this structure:

# Synthesis Report

**Generated:** [ISO 8601 timestamp]
**Research files analyzed:** [count]
**Total topics covered:** [count]

## Cross-cutting Themes

[For each theme that spans multiple topics:]

### [Theme Name]

- **Appears in:** [list of topic IDs]
- **Core concept:** [1-2 sentence explanation]
- **Why it matters:** [relevance to the domain]

## Knowledge Gaps

| Topic ID | Topic Title | Gap Description | Sources Consulted | Severity |
|----------|-------------|-----------------|-------------------|----------|
| ...      | ...         | ...             | [n], [m] from topic's research | high/med/low |

Note which source citations were checked when identifying each gap. If a gap exists because no source covered the subtopic, state "no sources found."

## Conflicting Information

### [Conflict Title]

- **Topics involved:** [topic IDs]
- **Source citations:** [cite the specific `[n]` references from each topic's research that conflict]
- **Conflict:** [description of the contradiction, referencing the cited sources]
- **Resolution:** [if known, how to reconcile; otherwise "needs investigation"]

## Topic Relationships

### Dependency Graph

[Describe the dependency ordering. Which topics are foundational? Which
build on others? Use indented lists or ASCII-style tree notation.]

### Independent Topics

[List topics that can be approached in any order without prerequisites.]

## Priority Rankings

| Priority | Topic ID | Topic Title | Rationale |
|----------|----------|-------------|-----------|
| critical | ...      | ...         | ...       |
| high     | ...      | ...         | ...       |
| medium   | ...      | ...         | ...       |
| low      | ...      | ...         | ...       |

## Recommended Order

[Numbered list of topics in the recommended sequence, considering
dependencies and priorities.]

Guidelines

• Be honest about gaps. It is more valuable to flag missing knowledge than to pretend coverage is complete.
• Prioritize actionable insights. Each section should help the reader decide what to focus on next.
• The dependency graph should be practical, not theoretical. Only list dependencies that actually affect comprehension.
• Keep the synthesis focused. This is a decision-making document, not a content summary.