Skip to content
Back to search

data2story-skill

Data Journalist Agent: Transforming Data into Verifiable Multimodal Story

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/qinghonglin--data2story-skill/badge.svg)](/listings/qinghonglin--data2story-skill)

HTML

<a href="/listings/qinghonglin--data2story-skill"><img src="/api/listings/qinghonglin--data2story-skill/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

Data2Story is an agent skill. The orchestrator lives in `skills/data2story/SKILL.md` — it works first-class with Claude Code, and equally with Codex, Cursor, Gemini CLI, and other agents.

1. Set your API key. Media generation routes through [OpenRouter](https://openrouter.ai/) by default:

   export OPENROUTER_API_KEY=sk-or-...

2. Run the skill on a dataset:

- **Claude Code** — make the skill available (place `skills/data2story/` under `~/.claude/skills/`, or run from inside this repo), then:

``` /data2story data/pick_a_card ```

- **Codex / other agents** — open the repo and ask the agent to follow the orchestrator:

``` Read skills/data2story/SKILL.md and run the Data2Story pipeline on data/pick_a_card ```

3. Open the output: `index.html` (the finished article) and `viewer.html` (the evidence inspector).

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 36 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. {
  2. "name": "data2story-skill",
  3. "owner": {
  4. "name": "QinghongLin",
  5. "url": "https://github.com/QinghongLin/data2story-skill"
  6. },
  7. "metadata": {
  8. "description": "Data Journalist Agent (Data2Story) skills — the paper's canonical pipeline and the extended pro build",
  9. "version": "1.0.0"
  10. },
  11. "plugins": [
  12. {
  13. "name": "data2story",
  14. "description": "The paper's canonical 7-role Data Journalist Agent pipeline",
  15. "source": "./",
  16. "strict": false,
  17. "skills": [
  18. "./skills/data2story",
  19. "./skills/frontend-design"
  20. ]
  21. },
  22. {
  23. "name": "data2story-pro",
  24. "description": "14-agent extended build — verified media, interactive playgrounds, cinematic cover, runnable verify layer",
  25. "source": "./",
  26. "strict": false,
  27. "skills": [
  28. "./skills/data2story-pro",
  29. "./skills/frontend-design-pro",
  30. "./skills/dataviz-craft",
  31. "./skills/find-data",
  32. "./skills/sparring-partner"
  33. ]
  34. }
  35. ]
  36. }
README.md · README · 128 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. # Data2Story
  2. A data-journalist agent skill that turns any dataset into a verifiable, evidence-grounded multimodal story — a self-contained HTML article where every sentence traces back to the data or source that justifies it.
  3. [![Website](https://img.shields.io/badge/🌐_Website-1a73e8)](https://data2story.github.io/)
  4. [![arXiv](https://img.shields.io/badge/arXiv-2606.11176-b31b1b?logo=arxiv&logoColor=white)](https://arxiv.org/abs/2606.11176)
  5. [![License: MIT](https://img.shields.io/badge/License-MIT-3da639.svg)](LICENSE)
  6. ## What it does
  7. https://github.com/user-attachments/assets/7a2e2b65-3289-431d-b06a-230252df8774
  8. - **Dataset → finished article.** Point it at a CSV, a folder of data, or a paper, and a fixed pipeline of seven roles produces a publishable HTML story end to end.
  9. - **Evidence-grounded.** Every sentence and visual links back to its source. The pipeline emits a `viewer.html` inspector where you can click any claim to see the data, code, or citation behind it.
  10. - **Multimodal by default.** Charts, images, video, audio, maps, and interactive elements — chosen from the data's actual properties, not a fixed checklist. Media generation routes through OpenRouter.
  11. - **Verifiable.** Each run writes to its own versioned project folder and snapshots the exact skill versions used, so every result can be traced and re-checked.
  12. - **Progressive disclosure.** Each role's `SKILL.md` holds only its instructions; bulky reference material (output schemas, field rules, lookup tables) lives in that role's `references/` folder as JSON and is loaded only when needed.
  13. > **Two versions.** The repository ships two peer skills under [`skills/`](skills/) — the richer, current build is **`data2story-pro`**; the paper's reproducible version is **`data2story`**.
  14. >
  15. > | Skill | Invoke | What it is |
  16. > |---|---|---|
  17. > | [`data2story-pro`](skills/data2story-pro/) | `/data2story-pro` | Improved version of data2story. See [`skills/data2story-pro/README.md`](skills/data2story-pro/README.md). |
  18. > | [`data2story`](skills/data2story/) | `/data2story` | The paper's canonical 7-role pipeline. |
  19. ## Updates
  20. * [x] [2026.6] Version 0.1.0: Optimized visual effects.
  21. * [x] [2026.6.9] Released the arXiv paper, GitHub repository.
  22. * [x] [2026.5.11] Data2Story was accepted as a Spotlight paper at the [ACM Conference on AI and Agentic Systems workshop](https://www.caisconf.org/).
  23. ## Installation & usage
  24. Data2Story is an agent skill. The orchestrator lives in `skills/data2story/SKILL.md` — it works first-class with Claude Code, and equally with Codex, Cursor, Gemini CLI, and other agents.
  25. 1. Set your API key. Media generation routes through [OpenRouter](https://openrouter.ai/) by default:
  26. ```bash
  27. export OPENROUTER_API_KEY=sk-or-...
  28. ```
  29. 2. Run the skill on a dataset:
  30. - **Claude Code** — make the skill available (place `skills/data2story/` under `~/.claude/skills/`, or run from inside this repo), then:
  31. ```
  32. /data2story data/pick_a_card
  33. ```
  34. - **Codex / other agents** — open the repo and ask the agent to follow the orchestrator:
  35. ```
  36. Read skills/data2story/SKILL.md and run the Data2Story pipeline on data/pick_a_card
  37. ```
  38. 3. Open the output: `index.html` (the finished article) and `viewer.html` (the evidence inspector).
  39. ## Project structure
  40. ```
  41. data2story-skill/
  42. ├── skills/
  43. │ ├── data2story/ the paper's canonical 7-role pipeline · /data2story
  44. │ ├── data2story-pro/ the extended 14-agent build · /data2story-pro
  45. │ │ └ self-contained: its own README, package.json, example data/
  46. │ ├── frontend-design/ visual design system used by data2story
  47. │ ├── frontend-design-pro/ extended design system used by data2story-pro
  48. │ ├── dataviz-craft/ chart-craft recipes · data2story-pro helper
  49. │ ├── find-data/ dataset discovery & validation · data2story-pro helper
  50. │ └── sparring-partner/ idea-to-topic ideation · data2story-pro helper
  51. ├── .claude-plugin/ marketplace.json · installs either skill
  52. └── assets/ shared images
  53. ```
  54. Inside each orchestrator, a role is a `SKILL.md` + a `references/` folder (JSON schemas & lookup tables) + any `scripts/` it runs — e.g. `designer/scripts/` (OpenRouter media tools), `inspector/scripts/` (the `verify.py` / `generate_viewer.py` verifier), `detective/scripts/` (media fetch helpers).
  55. ## The virtual newsroom
  56. Think of it as a small newsroom in a box. Each role reads what the previous one produced, then adds its own artifact — a fixed pipeline that runs once, end to end.
  57. | # | Role | What it does | Produces |
  58. |---|------|--------------|----------|
  59. | 1 | **Detective** | Researches external context — domain background, history, why the data matters | `detective.json` |
  60. | 2 | **Analyst** | Exhaustively profiles the data — distributions, correlations, trends, anomalies | `analyst.json`, `code/*.py` |
  61. | 3 | **Editor** | Decides the narrative — what the article argues and which findings matter | `editor.md`, `editor.json` |
  62. | 4 | **Designer** | Chooses how to show each point — charts, images, video, audio, interactives | `designer.json`, `assets/` |
  63. | 5 | **Programmer** | Builds the final HTML, tagging every element with its source IDs | `index.html` |
  64. | 6 | **Auditor** | Fixes layout issues — overlap, spacing, alignment — without changing content | `index.html` (fixed), `auditor.json` |
  65. | 7 | **Inspector** | Verifies every sentence traces to its evidence; builds an interactive viewer | `inspector.json`, `viewer.html` |
  66. ```mermaid
  67. flowchart TB
  68. DATA[(dataset)]
  69. subgraph plan [Plan and design]
  70. direction LR
  71. DET[Detective] -->|detective.json| ANA[Analyst]
  72. ANA -->|analyst.json + code| EDT[Editor]
  73. EDT -->|editor.md + editor.json| DES[Designer]
  74. end
  75. subgraph build [Build and verify]
  76. direction LR
  77. PRG[Programmer] -->|index.html| AUD[Auditor]
  78. AUD -->|index.html| INS[Inspector]
  79. end
  80. DATA --> DET
  81. DES -->|designer.json + assets| PRG
  82. INS -->|inspector.json + viewer.html| OUT([article + evidence viewer])
  83. ```
  84. ## License
  85. Released under the [MIT License](LICENSE).
  86. ## Acknowledgement
  87. If you use Data2Story in your research, please kindly cite:
  88. ```bibtex
  89. @article{data2story,
  90. title = {Data Journalist Agent: Transforming Data into Verifiable Multimodal Stories},
  91. author = {Lin, Kevin Qinghong and EI, Batu and Shi, Yuhong and Lu, Pan and Torr, Philip and Zou, James},
  92. journal = {arXiv preprint arXiv:2606.11176},
  93. year = {2026}
  94. }
  95. ```

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.