Skip to content
Back to search

Oh-my--paper

A Claude Code plugin that turns your terminal into an autonomous research lab — literature survey, experiment execution, paper writing, all in one pipeline.

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/ligphidonk--oh-my-paper/badge.svg)](/listings/ligphidonk--oh-my-paper)

HTML

<a href="/listings/ligphidonk--oh-my-paper"><img src="/api/listings/ligphidonk--oh-my-paper/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.

Step 1: Add the marketplace

/plugin marketplace add LigphiDonk/Oh-my--paper

Step 2: Install the plugin

/plugin install omp@oh-my-paper

Step 3: Restart Claude Code

Required for hooks to activate.

Step 4: Initialize your project

/omp:setup

This scaffolds the `.pipeline/` directory and registers the `SessionStart` hook for your project.

Update

The most reliable way to get the latest version:

/plugin uninstall omp
/plugin install omp@oh-my-paper
/reload-plugins

Or overwrite the plugin cache directly (faster, no restart needed):

cp -r /path/to/oh-my-paper/plugins/oh-my-paper/. \
  ~/.claude/plugins/cache/oh-my-paper/omp/1.0.0/
# Then in Claude Code:
/reload-plugins

Install from Local Directory

git clone https://github.com/LigphiDonk/Oh-my--paper.git /tmp/oh-my-paper
# In Claude Code:
/plugin marketplace add /tmp/oh-my-paper
/plugin install omp@oh-my-paper

---

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 · 21 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": "oh-my-paper",
  3. "owner": {
  4. "name": "LigphiDonk"
  5. },
  6. "metadata": {
  7. "description": "AI-powered academic research harness for Claude Code. Orchestrate literature surveys, idea generation, experiments, and paper writing with Codex delegation.",
  8. "version": "1.0.0"
  9. },
  10. "plugins": [
  11. {
  12. "name": "omp",
  13. "description": "Research harness: delegate to Codex, track memory across sessions, drive the full survey→ideation→experiment→publication pipeline.",
  14. "version": "1.0.0",
  15. "author": {
  16. "name": "LigphiDonk"
  17. },
  18. "source": "./plugins/oh-my-paper"
  19. }
  20. ]
  21. }
README.md · README · 495 lines

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

  1. <p align="center">
  2. <img src="./src/assets/qrcode.jpg" alt="交流群二维码" width="180" />
  3. <br/>
  4. <em>扫码加入交流群</em>
  5. </p>
  6. <p align="center">
  7. <img src="./icons/icon.png" alt="Oh My Paper" width="120" height="120" />
  8. </p>
  9. <h1 align="center">Oh My Paper</h1>
  10. <p align="center">
  11. <strong>A research harness for Claude Code — turn your terminal into an autonomous research lab.</strong>
  12. </p>
  13. <p align="center">
  14. <a href="./README.zh.md">中文文档</a>
  15. </p>
  16. <p align="center">
  17. <img src="https://img.shields.io/badge/claude--code-plugin-blueviolet?style=flat-square" />
  18. <img src="https://img.shields.io/badge/agents-5-ff69b4?style=flat-square" />
  19. <img src="https://img.shields.io/badge/skills-34-green?style=flat-square" />
  20. <img src="https://img.shields.io/badge/commands-8-blue?style=flat-square" />
  21. <img src="https://img.shields.io/badge/license-MIT-orange?style=flat-square" />
  22. </p>
  23. ---
  24. ## TL;DR
  25. ```bash
  26. # In Claude Code:
  27. /plugin marketplace add LigphiDonk/Oh-my--paper
  28. /plugin install omp@oh-my-paper
  29. ```
  30. Restart Claude Code. Run `/omp:setup` inside your research project, then drive the full pipeline with `/omp:survey`, `/omp:experiment`, and `/omp:write`. No GUI, no window-switching — everything in the terminal.
  31. ---
  32. ## Table of Contents
  33. - [Why This Exists](#why-this-exists)
  34. - [Install](#install)
  35. - [Claude Code Slash Commands](#claude-code-slash-commands)
  36. - [The Agent Team](#the-agent-team)
  37. - [34 Research Skills](#34-research-skills)
  38. - [Hooks](#hooks)
  39. - [Research Pipeline](#research-pipeline)
  40. - [Project Scaffold](#project-scaffold)
  41. - [How Memory Works](#how-memory-works)
  42. - [Codex Delegation](#codex-delegation)
  43. - [Remote Experiments](#remote-experiments)
  44. - [For LLM Agents](#for-llm-agents)
  45. - [Philosophy](#philosophy)
  46. - [Contributing](#contributing)
  47. - [Uninstall](#uninstall)
  48. ---
  49. ## Why This Exists
  50. Claude Code is already a great coding agent. But **research isn't just coding** — it's literature survey, idea evaluation, experiment design, paper writing, reference checking, and a dozen other things that require domain-specific workflows.
  51. Oh My Paper makes Claude Code **research-aware** by adding:
  52. - **A structured 5-stage pipeline** — Survey → Ideation → Experiment → Publication → Promotion
  53. - **5 specialized agent roles** — each with isolated memory and clear responsibilities
  54. - **34 built-in research skills** — from paper search to figure generation
  55. - **Background hooks** — auto-inject project context at session start, prompt role selection, track task completion
  56. - **Codex delegation** — hand off parallel tasks to Codex in a separate terminal
  57. Install it and forget about it. Your sessions get smarter. Your research gets organized.
  58. ---
  59. ## Install
  60. ### Step 1: Add the marketplace
  61. ```bash
  62. /plugin marketplace add LigphiDonk/Oh-my--paper
  63. ```
  64. ### Step 2: Install the plugin
  65. ```bash
  66. /plugin install omp@oh-my-paper
  67. ```
  68. ### Step 3: Restart Claude Code
  69. Required for hooks to activate.
  70. ### Step 4: Initialize your project
  71. ```bash
  72. /omp:setup
  73. ```
  74. This scaffolds the `.pipeline/` directory and registers the `SessionStart` hook for your project.
  75. ### Update
  76. The most reliable way to get the latest version:
  77. ```bash
  78. /plugin uninstall omp
  79. /plugin install omp@oh-my-paper
  80. /reload-plugins
  81. ```
  82. Or overwrite the plugin cache directly (faster, no restart needed):
  83. ```bash
  84. cp -r /path/to/oh-my-paper/plugins/oh-my-paper/. \
  85. ~/.claude/plugins/cache/oh-my-paper/omp/1.0.0/
  86. # Then in Claude Code:
  87. /reload-plugins
  88. ```
  89. ### Install from Local Directory
  90. ```bash
  91. git clone https://github.com/LigphiDonk/Oh-my--paper.git /tmp/oh-my-paper
  92. # In Claude Code:
  93. /plugin marketplace add /tmp/oh-my-paper
  94. /plugin install omp@oh-my-paper
  95. ```
  96. ---
  97. ## Claude Code Slash Commands
  98. These slash commands are provided by the **Claude Code plugin**.
  99. The Codex plugin does **not** currently auto-register `/omp-*` commands in the Codex CLI.
  100. All commands are prefixed with `/omp:`.
  101. | Command | What It Does |
  102. |---------|-------------|
  103. | `/omp:setup` | Scaffold a new research project — creates `.pipeline/`, memory files, and registers the SessionStart hook |
  104. | `/omp:survey` | AI-assisted literature survey — search papers, build `literature_bank.md` |
  105. | `/omp:ideate` | Generate and evaluate research ideas based on survey findings |
  106. | `/omp:experiment` | Design experiments, write evaluation code, run on remote compute nodes |
  107. | `/omp:write` | Draft paper sections, generate figures and captions, manage LaTeX files |
  108. | `/omp:review` | Peer-review your paper or experiment results before submission |
  109. | `/omp:delegate` | Generate a Codex prompt for a coding/experiment task; wait for result and update project state |
  110. | `/omp:plan` | Review global progress, confirm next steps, update research plan |
  111. ### Quick Start
  112. ```bash
  113. /omp:setup # scaffold the project
  114. /omp:survey # start literature survey
  115. /omp:ideate # generate ideas from survey
  116. /omp:experiment # design & run experiments
  117. /omp:write # draft the paper
  118. /omp:review # final quality gate
  119. ```
  120. ---
  121. ## The Agent Team
  122. When you open Claude Code in an Oh My Paper project, the `SessionStart` hook fires and Claude immediately asks which role you want to take on. Each role has **isolated memory** — it only reads and writes the files it needs.
  123. | Role | Responsibility | Memory Scope |
  124. |------|---------------|-------------|
  125. | **Conductor** | Global planning, review outputs, dispatch tasks, auto-update `project_truth` after each subtask | `project_truth` · `orchestrator_state` · `tasks.json` · `review_log` · `agent_handoff` · `decision_log` |
  126. | **Literature Scout** | Search papers, organize literature bank | `project_truth` · `execution_context` · `literature_bank` · `decision_log` |
  127. | **Experiment Driver** | Design experiments, write code, run evaluations | `execution_context` · `experiment_ledger` · `research_brief.json` · `project_truth` |
  128. | **Paper Writer** | Draft sections, generate figures, audit references | `execution_context` · `result_summary` · `literature_bank` · `agent_handoff` |
  129. | **Reviewer** | Peer review, quality gate, consistency check | `execution_context` · `project_truth` · `result_summary` |
  130. ### How It Works
  131. ```
  132. Session opens
  133. → SessionStart hook fires
  134. → Claude asks: which role today?
  135. → Agent loads role-specific memory files
  136. → Works as that persona
  137. → On subtask complete: auto-updates tasks.json + project_truth
  138. → Next session picks up right where you left off
  139. ```
  140. **Key design decisions:**
  141. - **Memory isolation** — the Paper Writer can't see the Conductor's orchestrator state; the Literature Scout can't see experiment results. This prevents context pollution.
  142. - **Shared state** — `tasks.json` and `project_truth.md` are the common ground, updated by all roles after each subtask.
  143. - **No manual sync** — the Conductor auto-updates `tasks.json` (marks tasks `done`) and appends a progress entry to `project_truth.md` whenever a subtask completes, without waiting for you to ask.
  144. ---
  145. ## 34 Research Skills
  146. Skills are structured instruction sets that Claude loads on demand. Each skill is a markdown file covering a specific research task.
  147. <details>
  148. <summary><strong>Click to expand the full skill list</strong></summary>
  149. | Category | Skills |
  150. |----------|--------|
  151. | **Literature** | `paper-finder` · `paper-analyzer` · `paper-image-extractor` · `research-literature-trace` · `biorxiv-database` · `dataset-discovery` |
  152. | **Survey & Ideation** | `inno-deep-research` · `gemini-deep-research` · `inno-code-survey` · `inno-idea-generation` · `inno-idea-eval` · `research-idea-convergence` |
  153. | **Experiment** | `inno-experiment-dev` · `inno-experiment-analysis` · `research-experiment-driver` · `remote-experiment` |
  154. | **Writing** | `inno-paper-writing` · `ml-paper-writing` · `scientific-writing` · `inno-figure-gen` · `inno-reference-audit` · `research-paper-handoff` |
  155. | **Planning & Review** | `inno-pipeline-planner` · `research-pipeline-planner` · `inno-paper-reviewer` · `inno-prepare-resources` · `inno-rclone-to-overleaf` |
  156. | **Presentation** | `making-academic-presentations` · `inno-grant-proposal` |
  157. | **Agent Dispatch** | `claude-code-dispatch` · `codex-dispatch` |
  158. | **Domain-Specific** | `academic-researcher` · `bioinformatics-init-analysis` · `research-news` |
  159. </details>
  160. Skills are auto-recommended based on your current pipeline stage. Add project-local skills in the `skills/` directory.
  161. ---
  162. ## Hooks
  163. Oh My Paper registers three hooks that run in the background:
  164. | Hook | Trigger | What It Does |
  165. |------|---------|-------------|
  166. | **SessionStart** | Every time you open Claude Code in this project | Outputs project context to Claude — current stage, active task, last handoff — then prompts you to pick a role via `AskUserQuestion` |
  167. | **Stop** | When a task completes | Tracks task completion, updates `tasks.json` |
  168. | **PostToolUse (Write)** | After any file write | Detects pipeline stage transitions |
  169. **Important:** Hooks only activate after running `/omp:setup` in your project. Setup registers the `SessionStart` hook in `.claude/settings.json` and creates the `.pipeline/` directory that the hook checks.
  170. ---
  171. ## Research Pipeline
  172. A structured 5-stage workflow from idea to publication:
  173. ```
  174. ┌──────────┐ ┌──────────┐ ┌────────────┐ ┌─────────────┐ ┌───────────┐
  175. │ Survey │ → │ Ideation │ → │ Experiment │ → │ Publication │ → │ Promotion │
  176. └──────────┘ └──────────┘ └────────────┘ └─────────────┘ └───────────┘
  177. ```
  178. Each stage comes with:
  179. - **Auto-generated task trees** — what to do next
  180. - **Recommended skills** — which skills to load
  181. - **Context-aware prompts** — agents read `tasks.json` and `research_brief.json` and know what to do
  182. ---
  183. ## Project Scaffold
  184. `/omp:setup` creates this structure:
  185. ```
  186. my-research/
  187. ├── paper/ # LaTeX workspace
  188. │ ├── main.tex
  189. │ ├── sections/
  190. │ └── refs/
  191. ├── experiment/ # Experiment code & scripts
  192. ├── survey/ # Literature survey artifacts
  193. ├── ideation/ # Ideas, evaluations, plans
  194. ├── promotion/ # Slides, demos, outreach
  195. ├── skills/ # Project-local skills
  196. ├── .pipeline/
  197. │ ├── tasks/
  198. │ │ └── tasks.json # Task tree across all stages
  199. │ ├── docs/
  200. │ │ └── research_brief.json
  201. │ └── memory/ # Agent memory files
  202. ├── .claude/
  203. │ └── settings.json # SessionStart hook registration
  204. ├── CLAUDE.md
  205. └── AGENTS.md
  206. ```
  207. ---
  208. ## How Memory Works
  209. Each agent role reads and writes specific memory files. The Conductor is responsible for keeping shared state in sync.
  210. ```
  211. .pipeline/memory/
  212. ├── project_truth.md # Ground truth + progress log (appended after each subtask)
  213. ├── orchestrator_state.md # Conductor's planning state
  214. ├── execution_context.md # Current task context for executors
  215. ├── experiment_ledger.md # Experiment history & results
  216. ├── result_summary.md # Latest results for writing & review
  217. ├── review_log.md # Review feedback history
  218. ├── literature_bank.md # Organized paper notes
  219. ├── agent_handoff.md # Cross-agent handoff messages
  220. └── decision_log.md # Rejected directions & reasoning
  221. .pipeline/tasks/
  222. └── tasks.json # Shared task tree (all roles read/write this)
  223. ```
  224. Memory survives across sessions. The `SessionStart` hook reads these files and injects the relevant context — you pick up right where you left off.
  225. **Auto-sync rule:** The Conductor updates `tasks.json` and `project_truth.md` automatically after every subtask completes (delegate / experiment / survey / write / review). You never need to ask it to sync.
  226. ---
  227. ## Codex Delegation
  228. The Conductor can hand off coding and experiment tasks to Codex:
  229. ```bash
  230. /omp:delegate
  231. ```
  232. The flow:
  233. 1. Conductor reads project context and the current task
  234. 2. Presents task summary — you confirm
  235. 3. Generates a complete Codex prompt with context pre-injected
  236. 4. You copy it to a new terminal: `codex "..."`
  237. 5. Conductor polls for completion (`CODEX_DONE` signal in `agent_handoff.md`)
  238. 6. Reads result, asks you to accept/revise/reject
  239. 7. On accept: updates `tasks.json` and `project_truth.md` automatically
  240. ---
  241. ## Remote Experiments
  242. The `remote-experiment` skill + `/omp:experiment` support a full auto-experiment loop:
  243. ```
  244. Design plan → Implement code → rsync to server → Run on GPU → Parse metrics → Repeat
  245. ```
  246. - SSH/rsync-based remote compute via `compute-helper` CLI
  247. - Configurable success thresholds, max iterations, and failure limits
  248. - Results flow back into `experiment_ledger.md` for the Paper Writer
  249. ---
  250. ## For LLM Agents
  251. If you're an AI agent installing this plugin:
  252. ```bash
  253. # Step 1: Add marketplace
  254. /plugin marketplace add LigphiDonk/Oh-my--paper
  255. # Step 2: Install plugin
  256. /plugin install omp@oh-my-paper
  257. # Step 3: Verify installation
  258. /plugin
  259. # Should show: omp @ oh-my-paper, Status: Enabled
  260. # Step 4: User must restart Claude Code (you cannot do this)
  261. # Tell user: "Please restart Claude Code to activate hooks."
  262. # Step 5: Initialize project
  263. /omp:setup
  264. ```
  265. ---
  266. ## Philosophy
  267. > **Enhance, don't replace.** Claude Code is already smart — we add research structure, not overrides.
  268. - **Your context is for reasoning** — hooks inject only what's needed; memory files keep the rest on disk
  269. - **Domain-specific, not generic** — every skill, agent, and command is designed for academic research
  270. - **Invisible when not needed** — hooks run in the background; no noise if you're just coding
  271. - **Composable** — use one command, use all of them, or just let the hooks do their thing
  272. - **Memory over repetition** — agents remember project context so you don't re-explain every session
  273. ---
  274. ## Contributing
  275. PRs welcome. If you add a new skill, put it in `skills/` with proper YAML frontmatter and update `research-catalog.json`.
  276. Any change to cached content requires version bumps in **both**:
  277. - `plugins/oh-my-paper/.claude-plugin/plugin.json`
  278. - `.claude-plugin/marketplace.json`
  279. ---
  280. ## Codex Support
  281. Oh My Paper also ships a **Codex plugin** (`oh-my-paper-codex`) that shares the same research harness concepts, agents, and skills as the Claude Code plugin.
  282. ### Install on Codex
  283. **macOS / Linux**
  284. ```bash
  285. # 1. Clone the repo
  286. git clone https://github.com/LigphiDonk/Oh-my--paper.git /tmp/oh-my-paper
  287. cd /tmp/oh-my-paper
  288. # 2. One-command install
  289. ./scripts/install-codex-plugin.sh
  290. ```
  291. **Windows (PowerShell)**
  292. ```powershell
  293. # 1. Clone the repo
  294. git clone https://github.com/LigphiDonk/Oh-my--paper.git $env:TEMP\oh-my-paper
  295. Set-Location $env:TEMP\oh-my-paper
  296. # 2. One-command install
  297. powershell -ExecutionPolicy Bypass -File .\scripts\install-codex-plugin.ps1
  298. ```
  299. What the installer does:
  300. - Copies the plugin to `~/plugins/oh-my-paper-codex`
  301. - Creates or updates `~/.agents/plugins/marketplace.json`
  302. - Tries to call Codex directly so the plugin becomes installed and enabled immediately
  303. - Uses `node` under the hood, so make sure `node` is available on your `PATH`
  304. If `codex` is not available on your `PATH`, the script still registers the plugin and then tells you to finish the last step in Codex's Plugins page. If you search there, search for `Oh My Paper` or `oh-my-paper-codex`, not `omp`.
  305. ### Use in Codex CLI
  306. After installation, start Codex in your research project directory:
  307. ```bash
  308. cd /path/to/your/research-project
  309. codex
  310. ```
  311. Then use one of these two patterns:
  312. - Ask naturally, for example: `Use Oh My Paper to initialize this research project and scaffold .pipeline/`
  313. - Reuse the workflow prompt templates under `plugins/oh-my-paper-codex/prompts/` by copying or adapting them inside the Codex session
  314. Codex CLI does **not** currently auto-register the files in `plugins/oh-my-paper-codex/prompts/` as slash commands, so `/omp-setup` and similar commands will **not** appear in the CLI command palette.
  315. ### What's Included
  316. | Feature | Claude Code | Codex CLI |
  317. |:---|:---|:---|
  318. | Agent Roles (5) | `agents/*.md` | `agents/*.toml` |
  319. | Workflow entrypoints | `/omp:...` slash commands | Natural-language prompts + `prompts/*.md` templates |
  320. | SessionStart Hook | Native hook | `AGENTS.md` (auto-read) |
  321. | Skills (34) | ✅ shared | ✅ shared |
  322. | `.pipeline/` Memory | ✅ | ✅ |
  323. | Codex Delegation | `/omp:delegate` → new terminal | Native `/agent` subagent |
  324. ### Key Differences
  325. - **Hooks**: Codex doesn't have native hooks. The `SessionStart` equivalent is handled by `AGENTS.md` which Codex reads automatically. Stage transition detection is embedded in the agent instructions.
  326. - **CLI command model**: Claude Code exposes `/omp:...` slash commands. Codex CLI currently does not auto-register the plugin's `prompts/*.md` files as `/omp-*` slash commands, so you use natural-language prompts or copy/adapt the templates manually.
  327. - **Both can coexist**: The Codex plugin (`plugins/oh-my-paper-codex/`) is completely separate from the Claude Code plugin (`plugins/oh-my-paper/`). Installing one does not affect the other.
  328. - **Installer scripts**: Use `scripts/install-codex-plugin.sh` on macOS/Linux or `scripts/install-codex-plugin.ps1` on Windows. They merge the marketplace entry instead of overwriting your existing local plugins.
  329. - **Codex discovery**: Codex expects a valid `~/.agents/plugins/marketplace.json` entry plus a plugin directory under `~/plugins/<plugin-name>/`. Copying files only into `~/.codex/plugins/` is not enough for the plugin UI to discover it.
  330. - **Codex install state**: A marketplace entry only makes the plugin appear in the Plugins page. You must still install it there before it becomes enabled and usable.
  331. ---
  332. ## Uninstall
  333. **Claude Code:**
  334. ```bash
  335. /plugin uninstall omp@oh-my-paper
  336. ```
  337. **Codex on macOS / Linux:**
  338. ```bash
  339. ./scripts/uninstall-codex-plugin.sh
  340. ```
  341. **Codex on Windows (PowerShell):**
  342. ```powershell
  343. powershell -ExecutionPolicy Bypass -File .\scripts\uninstall-codex-plugin.ps1
  344. ```
  345. ---
  346. ## License
  347. MIT. See [LICENSE](./LICENSE).
  348. ---
  349. ## Acknowledgments
  350. Special thanks to the **[Linux.do](https://linux.do)** community for your support and feedback.
  351. ---
  352. <p align="center">
  353. <strong>Oh My Paper</strong> — Where Research Meets the Terminal.
  354. </p>

Ratings and comments

Sign in to rate this skill.

Sign in

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