# Comparative Synthesis

> Compare and synthesize findings across multiple completed DeepScan reports. Use when the user wants cross-run analysis, trend comparison, or a unified summary from several research sessions.

- **Type:** Skill
- **Install:** `agentstack add skill-papersflow-ai-papersflow-skills-comparative-synthesis`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [papersflow-ai](https://agentstack.voostack.com/s/papersflow-ai)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [papersflow-ai](https://github.com/papersflow-ai)
- **Source:** https://github.com/papersflow-ai/papersflow-skills/tree/main/skills/comparative-synthesis

## Install

```sh
agentstack add skill-papersflow-ai-papersflow-skills-comparative-synthesis
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Comparative Synthesis

Use this skill when the user wants to compare, contrast, or synthesize findings across multiple completed DeepScan runs rather than monitor a single active job.

## Workflow

1. Use `summarize_evidence` to pull cross-report summaries from the user's DeepScan history.
2. If the user references specific runs, use `get_deepscan_report` for each to get full report data.
3. Identify overlapping papers, conflicting findings, and complementary themes across runs.
4. Use `run_python_plot` to visualize comparisons when the data supports it.

## Output Style

Structure the synthesis around:

- **Common ground** — papers, methods, or findings that appear across multiple runs
- **Divergences** — where different runs reached different conclusions or surfaced different literature
- **Gaps** — topics or questions that no run adequately covered
- **Trends** — temporal patterns, emerging methods, or shifting consensus visible across runs

Keep sections short and reference specific papers by title and year.

## Tool Guidance

### Use `summarize_evidence`

Call this first. It aggregates across the user's stored DeepScan history and is the fastest way to get a cross-run view.

Use for:

- "What do my recent DeepScans say about X?"
- "Summarize everything I've researched on topic Y"
- "Compare findings across my last three runs"

### Use `get_deepscan_report`

Call for specific runs when the user wants:

- side-by-side comparison of two named runs
- detailed data from a particular session that `summarize_evidence` condensed too aggressively

### Use `run_python_plot`

Use after you have structured data from reports. Good comparison plots include:

- paper overlap Venn or bar chart across runs
- citation count distributions side by side
- publication year histograms per run
- venue frequency comparison
- topic/method co-occurrence heatmap

Only plot when there is enough data to be meaningful. Say so if the data is too sparse.

### Do NOT use

- `run_deepscan` — this skill synthesizes completed runs, not starts new ones
- `search_literature` — use the existing DeepScan data, not new searches

## Examples

- User asks: "Compare my DeepScan on transformer efficiency with the one on model distillation."
- User asks: "What themes keep showing up across all my recent research sessions?"
- User asks: "Plot the publication year distribution from my last two DeepScans side by side."
- User asks: "Synthesize everything I've researched on protein folding this month."

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [papersflow-ai](https://github.com/papersflow-ai)
- **Source:** [papersflow-ai/papersflow-skills](https://github.com/papersflow-ai/papersflow-skills)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-papersflow-ai-papersflow-skills-comparative-synthesis
- Seller: https://agentstack.voostack.com/s/papersflow-ai
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.