Plugin

paper-wiki

Name: paper-wiki
Author: kaneyxx

By kaneyxx

Personal research wiki builder for Claude Code — pipeline-driven paper ingestion with plugin architecture.

Component Overview

/analyze, /bio-search +8

Commands

Agents

analyze, bio-search +8

Skills

SessionStart

Hooks

MCP Servers

LSP Servers

Output Styles

Themes

Monitors

Install

npx claudepluginhub kaneyxx/paper-wiki --plugin paper-wiki

Component Details

Commands (10)

Analyze

/analyze

Deep-analyze a single paper into a wiki entry

Bio Search

/bio-search

Search biomedical literature (bioRxiv, medRxiv, PubMed Central) via paperclip MCP

Digest

/digest

Build a paper-wiki digest from a recipe and write it to disk

Extract Images

/extract-images

Pull real paper figures from arXiv source and embed them in Wiki/sources/

Migrate Sources

/migrate-sources

Upgrade existing Wiki/sources/<id>.md files to the current section-organized format

Setup

/setup

Verify paper-wiki environment and configure first-time setup

Wiki Compile

/wiki-compile

Rebuild Wiki/index.md from current concept and source articles

Wiki Ingest

/wiki-ingest

Fold a source paper into the wiki, regenerating affected concept articles

Wiki Lint

/wiki-lint

Health-check the wiki and propose batch fixes

Wiki Query

/wiki-query

Answer a research question from the wiki with cited sources

Skills (10)

analyze

/analyze

Deep-analyzes a single paper into a wiki entry. Use when the user invokes /paperwiki:analyze, asks "tell me more about <paper>", wants a per-paper breakdown of methods and contributions, or is following up on a recommendation from /paperwiki:digest.

bio-search

/bio-search

Searches biomedical literature (bioRxiv, medRxiv, PubMed Central) via the paperclip MCP server and optionally folds results into the wiki. Use when the user invokes /paperwiki:bio-search, asks to "search bioRxiv", "search PubMed", "find biomedical papers on X", "look up a clinical trial on Y", or otherwise wants biomedical-domain literature beyond what arXiv covers.

digest

/digest

Builds a research-paper digest from a recipe, writing it to disk via the configured reporters. Use when the user asks for "today's papers", invokes /paperwiki:digest, requests a daily/weekly research roundup, or wants to refresh their paper-wiki vault.

extract-images

/extract-images

Pulls real paper figures from arXiv source tarballs and embeds them in the source's Wiki/sources/<id>.md file. Use when the user invokes /paperwiki:extract-images, asks to "show me the figures from <paper>", "pull the architecture diagram", "extract images from <id>", or follows up on a digest entry that has an empty Figures section.

migrate-sources

/migrate-sources

Upgrades existing Wiki/sources/<id>.md files to the current source-stub format while preserving user-edited content in Notes / Key Takeaways / Figures sections. Use when /paperwiki:migrate-sources is invoked, after upgrading paper-wiki to a new minor version, when wiki-lint surfaces stale-format warnings, or when the user notices their old source files lack the Core Information / Key Takeaways / Figures section structure.

setup

/setup

Interactive first-run wizard. Verifies paper-wiki's Python environment, walks the user through five questions to build their personal recipe (vault path, topics, S2 API key, auto-ingest preference), writes the config files, and surfaces optional MCP servers. Use when the user invokes /paperwiki:setup, when no personal recipe exists yet at ~/.config/paper-wiki/recipes/, or when downstream paperwiki SKILLs report missing config.

wiki-compile

/wiki-compile

Rebuilds Wiki/index.md deterministically from current concept and source articles. Use when /paperwiki:wiki-compile is invoked, after a wiki-ingest run, before sharing the wiki, or whenever the user wants the index refreshed.

wiki-ingest

/wiki-ingest

Ingests a new source paper into the wiki by regenerating affected concept articles. Use when /paperwiki:wiki-ingest is invoked, when a fresh source lands in Wiki/sources/ via analyze or digest, or when the user says "fold this paper into the wiki" or "update the wiki with X".

wiki-lint

/wiki-lint

Runs a health check on the wiki and surfaces orphans, stale entries, oversized files, broken wikilinks, and status mismatches. Use when /paperwiki:wiki-lint is invoked, before a release of the wiki to others, or periodically to keep concept articles trustworthy.

wiki-query

/wiki-query

Answers a research question by searching the wiki and synthesizing an answer with citations. Use when /paperwiki:wiki-query is invoked, when the user asks "what does my wiki say about X", or when the user wants a literature-aware answer drawn from previously-ingested sources.

Hooks (1)

Review workflow modifications before installing

Event Hooks

1 hook across 1 event

README

paper-wiki

A personal research wiki builder for Claude Code. Pipeline-driven paper ingestion with a plugin architecture, persistent knowledge-accumulating wiki, and zero LLM API keys (Claude Code is the LLM).

paper-wiki turns the firehose of academic publishing into a curated, queryable, compounding wiki. SKILLs let you fetch papers from arXiv, Semantic Scholar, and (optionally) paperclip's biomedical corpus; filter and score them against your research interests; persist durable notes; and let Claude synthesize topic articles that link sources together — all from inside Claude Code.

Prerequisites

Python ≥ 3.11. The plugin self-installs its .venv on first Claude Code session via hooks/ensure-env.sh. uv is preferred and used automatically when on PATH; otherwise python3 -m venv + pip.
Claude Code CLI ≥ 2.1.119 — paper-wiki is a Claude Code plugin; SKILLs run inside Claude Code, not standalone. v0.3.4's setup wizard depends on AskUserQuestion's header field which versions older than 2.1.119 do not render correctly (earlier versions auto-truncate the label).
Optional: a Semantic Scholar API key (free) — bumps the rate limit from ~1 req/s to 100 req/s and is strongly recommended for any non-trivial use.
Optional: paperclip for biomedical literature; see Optional: biomedical literature below for details on the CLI vs MCP paths.

Install

In Claude Code:

/plugin marketplace add kaneyxx/paper-wiki
/plugin install paper-wiki@paper-wiki

The plugin self-installs its Python environment on first session — no manual setup required.

Advanced: custom config location (dotfiles users)

By default paper-wiki stores your personal recipe and secrets under ~/.config/paper-wiki/. If you keep your config under a custom path (e.g. ~/dotfiles/paper-wiki/), set the PAPERWIKI_CONFIG_DIR environment variable before running the setup wizard and paper-wiki will use that directory instead. Resolution priority:

1. $PAPERWIKI_CONFIG_DIR          (if set)
2. $XDG_CONFIG_HOME/paper-wiki    (if XDG_CONFIG_HOME is set)
3. ~/.config/paper-wiki           (default)

First-run walkthrough

1. Install

/plugin marketplace add kaneyxx/paper-wiki
/plugin install paper-wiki@paper-wiki

2. Run the setup wizard

/paper-wiki:setup

The wizard asks 5–8 questions (vault path, topics, Semantic Scholar API key, auto-ingest depth, paperclip availability) and writes:

~/.config/paper-wiki/recipes/daily.yaml — your personal recipe
~/.config/paper-wiki/secrets.env — API keys (chmod 600)

Re-running the wizard against an existing config offers Keep / Reconfigure all / Edit one piece — non-destructive by default.

3. Generate your first digest

/paper-wiki:digest daily

Pulls papers from arXiv + Semantic Scholar, scores them against your topics, and writes ~/Documents/<your-vault>/Daily/<YYYY-MM-DD>.md with Obsidian callouts, inline teaser figures, and ### Detailed report sub-headings per paper.

If your recipe has auto_ingest_top: 3, the top 3 papers are automatically folded into the wiki concept articles via /paper-wiki:wiki-ingest at the end of the digest run.

4. Browse and query

/paper-wiki:wiki-query "vision-language pathology"

Keyword + tag search across Wiki/concepts/ and Wiki/sources/, with Claude synthesizing a cited answer.

Manual setup (fallback)

If you prefer to bypass the wizard and configure by hand, the recipes under recipes/*.yaml are templates. Copy one as a starting point:

mkdir -p ~/.config/paper-wiki/recipes
cp recipes/daily-arxiv.yaml ~/.config/paper-wiki/recipes/daily.yaml

Then edit ~/.config/paper-wiki/recipes/daily.yaml:

Change every vault_path and vault_paths reference to your vault
Change the topics to keywords you actually read about
Add api_key_env: PAPERWIKI_S2_API_KEY to the semantic_scholar source if you have a key (recommended)

Store the API key securely:

mkdir -p ~/.config/paper-wiki
cat > ~/.config/paper-wiki/secrets.env <<'EOF'
export PAPERWIKI_S2_API_KEY="<your-S2-key>"
EOF
chmod 600 ~/.config/paper-wiki/secrets.env

Bundled and personal recipes never inline the key. They reference it via api_key_env: PAPERWIKI_S2_API_KEY; the recipe loader resolves the env var at pipeline-build time, so the secret stays out of any YAML file.

The vault subdirectory layout paper-wiki creates by default:

~/Documents/Paper-Wiki/
├── Daily/      # daily digest output (one .md per day)
├── Sources/    # per-paper notes (created by /paper-wiki:analyze)
└── Wiki/       # synthesized concept articles + per-paper source stubs
    ├── sources/
    ├── concepts/
    └── index.md

View full README on GitHub

Similar Plugins

claude-obsidian

3.2k

Claude + Obsidian knowledge companion. Sets up a persistent, compounding wiki vault. Covers memory management, session notetaking, knowledge organization, and agent context across projects. Based on Andrej Karpathy's LLM Wiki pattern. Optional DragonScale Memory extension adds hierarchical log folds, deterministic page addresses, embedding-based semantic tiling lint, and boundary-first autoresearch topic selection.

Stats

Version0.3.5

Stars0

MaintenanceExcellent

LicenseGPL-3.0

AddedApr 26, 2026

Actions

View on GitHub View README Plugin Marketplace JSON

Available In

paper-wiki

A personal research wiki builder for Claude Code. Pipeline-driven paper ingestion with a plugin architecture, persistent knowledge-accumulating wiki, and zero LLM API keys (Claude Code is the LLM).

Prerequisites

Python ≥ 3.11. The plugin self-installs its .venv on first Claude Code session via hooks/ensure-env.sh. uv is preferred and used automatically when on PATH; otherwise python3 -m venv + pip.
Claude Code CLI ≥ 2.1.119 — paper-wiki is a Claude Code plugin; SKILLs run inside Claude Code, not standalone. v0.3.4's setup wizard depends on AskUserQuestion's header field which versions older than 2.1.119 do not render correctly (earlier versions auto-truncate the label).
Optional: a Semantic Scholar API key (free) — bumps the rate limit from ~1 req/s to 100 req/s and is strongly recommended for any non-trivial use.
Optional: paperclip for biomedical literature; see Optional: biomedical literature below for details on the CLI vs MCP paths.

Install

In Claude Code:

/plugin marketplace add kaneyxx/paper-wiki
/plugin install paper-wiki@paper-wiki

The plugin self-installs its Python environment on first session — no manual setup required.

Advanced: custom config location (dotfiles users)

1. $PAPERWIKI_CONFIG_DIR          (if set)
2. $XDG_CONFIG_HOME/paper-wiki    (if XDG_CONFIG_HOME is set)
3. ~/.config/paper-wiki           (default)

First-run walkthrough

1. Install

/plugin marketplace add kaneyxx/paper-wiki
/plugin install paper-wiki@paper-wiki

2. Run the setup wizard

/paper-wiki:setup

The wizard asks 5–8 questions (vault path, topics, Semantic Scholar API key, auto-ingest depth, paperclip availability) and writes:

~/.config/paper-wiki/recipes/daily.yaml — your personal recipe
~/.config/paper-wiki/secrets.env — API keys (chmod 600)

Re-running the wizard against an existing config offers Keep / Reconfigure all / Edit one piece — non-destructive by default.

3. Generate your first digest

/paper-wiki:digest daily

If your recipe has auto_ingest_top: 3, the top 3 papers are automatically folded into the wiki concept articles via /paper-wiki:wiki-ingest at the end of the digest run.

4. Browse and query

/paper-wiki:wiki-query "vision-language pathology"

Keyword + tag search across Wiki/concepts/ and Wiki/sources/, with Claude synthesizing a cited answer.

Manual setup (fallback)

If you prefer to bypass the wizard and configure by hand, the recipes under recipes/*.yaml are templates. Copy one as a starting point:

mkdir -p ~/.config/paper-wiki/recipes
cp recipes/daily-arxiv.yaml ~/.config/paper-wiki/recipes/daily.yaml

Then edit ~/.config/paper-wiki/recipes/daily.yaml:

Change every vault_path and vault_paths reference to your vault
Change the topics to keywords you actually read about
Add api_key_env: PAPERWIKI_S2_API_KEY to the semantic_scholar source if you have a key (recommended)

Store the API key securely:

mkdir -p ~/.config/paper-wiki
cat > ~/.config/paper-wiki/secrets.env <<'EOF'
export PAPERWIKI_S2_API_KEY="<your-S2-key>"
EOF
chmod 600 ~/.config/paper-wiki/secrets.env

The vault subdirectory layout paper-wiki creates by default:

~/Documents/Paper-Wiki/
├── Daily/      # daily digest output (one .md per day)
├── Sources/    # per-paper notes (created by /paper-wiki:analyze)
└── Wiki/       # synthesized concept articles + per-paper source stubs
    ├── sources/
    ├── concepts/
    └── index.md

paper-wiki

Component Overview

Install

Component Details

Commands (10)

Skills (10)

Hooks (1)

README

paper-wiki

Prerequisites

Install

Advanced: custom config location (dotfiles users)

First-run walkthrough

1. Install

2. Run the setup wizard

3. Generate your first digest

4. Browse and query

Manual setup (fallback)

Similar Plugins

claude-obsidian

paper-wiki

Component Overview

Install

Component Details

Commands (10)

Skills (10)

Hooks (1)

README

paper-wiki

Prerequisites

Install

Advanced: custom config location (dotfiles users)

First-run walkthrough

1. Install

2. Run the setup wizard

3. Generate your first digest

4. Browse and query

Manual setup (fallback)

Similar Plugins

claude-obsidian

vibeworks-library

context7-plugin

deep-wiki