SkillAgentSearch skills...

organize-wiki

Use when the user wants to organize their paper library (整理知识库): review or change its domain folders, write the domain convention DeepPaperNote follows, refile misplaced papers, fix papers/ tags, or undo the last reorganization.

Install / Use

npx skills add 917Dhj/DeepPaperNote --skill organize-wiki

Installs into whichever agent you are using.

About this skill
📄

SKILL.md

Installable skill definition

Quality Score

85/100

Supported Platforms

Universal

Our assessment of organize-wiki

organize-wiki scores 85/100 on our quality scale, 316th of 435 Education & Research skills we index.

Its SKILL.md is 4.9 KB long, split into 6 sections and no code examples: a solid amount of guidance for an agent.

With 1,112 GitHub stars, it is one of the more widely adopted skills in the catalogue.

Substance
26/30
Structure
11/20
Description
15/15
Adoption
13/20
Freshness
15/15

Maintenance, license and trust

  • The repository was last updated about 2 months ago, so organize-wiki is actively maintained.
  • It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
  • Its trust signals score 100/100, with no cautions. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.

organize-wiki compared with similar skills

All 4 of these similar skills score higher than organize-wiki; compare them before choosing.

SkillScoreStarsUpdatedFormat
organize-wiki (this skill)by 917Dhj851.1k48d agoSKILL.md
last30days-skillby mvanhorn10067.0ktodayCLAUDE.md
algorithmic-artby anthropics100177.9k47d agoSKILL.md
pptxby anthropics100177.9k47d agoSKILL.md
designby nextlevelbuilder100143.4k8d agoSKILL.md

Frequently asked questions

How do I install organize-wiki?
Run npx skills add 917Dhj/DeepPaperNote --skill organize-wiki. The install tabs above show the steps for each supported agent.
Which AI agents does organize-wiki work with?
It is written for Universal, as a SKILL.md file. Other agents that read the same format can often use it too.
Is organize-wiki safe to use?
It is MIT-licensed and scores 100/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
Is organize-wiki still maintained?
The repository was last updated about 2 months ago, so organize-wiki is actively maintained.

name: organize-wiki description: Use when the user wants to organize their paper library (整理知识库): review or change its domain folders, write the domain convention DeepPaperNote follows, refile misplaced papers, fix papers/ tags, or undo the last reorganization.

Organize Wiki

Keep the user's paper library in domain folders they agree with. You propose, with evidence from the library and their research memory; the user decides. Only a confirmed plan reaches the vault, only through scripts/organize.py, and only after the user has seen its preview.

references/file-contract.md defines the library layout, the domains file, the scan, the plan, and the log. Read it before step 3.

To undo the last reorganization, run python3 scripts/organize.py undo --latest, report what it restored or the conflicts it lists, and stop.

1. Read the library

Run python3 scripts/scan_library.py --output <tmp>/scan.json. It finds the vault through DeepPaperNote's configuration. On needs_vault or not_a_vault, ask the user once for their Obsidian vault folder and pass --vault <path> to every script.

Done when the scan reports status: ok.

2. Agree on the convention

The convention is the set of domains, each with a definition, what it includes, what it excludes and where that goes instead, and three to five representative papers. domains_file.exists picks the mode.

Building (no domains file): draft each existing domain's fields from its papers' titles and excerpts, then propose changes to the set:

  • merge a domain holding one or two papers, or two domains whose papers mostly overlap;
  • split a domain past about 25 papers that holds at least two clear sub-areas of five or more papers each;
  • rename a domain whose name is broader or vaguer than what its papers are about;
  • add a domain for papers no domain covers.

Updating (the file exists): every domain in it counts as confirmed. Ask only about change: papers no domain includes, papers a domain's own excludes rule out, domains grown past about 25 or shrunk to one or two, definitions that overlap.

In both modes:

  • With research_memory.exists, split finer along the user's directions and keep not-focus topics coarse. Without it, say once that a research memory would fit the split to their research, and go on.
  • The thresholds are guides; give the reason when you depart from them.
  • Domains stay one level: a split makes sibling folders. A new name is a specific research area at the level of its siblings, written with letters, digits, _, and - only.

Interview in rounds of at most four numbered questions in the user's language. Each asks one decision, with your proposal and its evidence (paper counts and titles), worded so "yes" accepts it. Show the full convention draft once the rounds settle it.

Done when the user has confirmed the complete convention.

3. Plan the changes

Assign every paper to a domain of the convention by its title and excerpt; its current folder is weak evidence. Present the plan in groups, each item numbered with a one-line reason:

  1. Folder changes: domains created, renamed, merged, or removed.
  2. Moves: each misfiled paper, from and to.
  3. Tag fixes: papers that stay put with tag_status mismatch or multiple. Moved papers get the same fix automatically. Notes without frontmatter stay untouched.
  4. Structure: flatten each nested_folder, wrap each loose_file in its own paper folder, and move each paper_at_root into a domain. Report each images_index without changing it.
  5. Needs your judgment: each paper you cannot place with confidence, asked as a question with your best guess.

The user accepts everything or rejects items by number. Write what they accepted as the plan JSON.

Done when every item is accepted or rejected and no question is open.

4. Preview, confirm, apply

Run python3 scripts/organize.py preview --plan <plan.json>. On refused, fix the plan for each error and rerun.

Show the user the summary counts, the moves, the domains file diff (the full text when it is new), and the warnings. Name any note outside the paper folder whose links will be rewritten.

After the user confirms, run python3 scripts/organize.py apply --plan <plan.json> --expected-digest <plan_digest>. A refusal because the library changed means rerunning preview and showing what differs before asking again. A lock refusal means DeepPaperNote is saving a note; retry once it finishes.

Done when apply reports status: ok with verify.status: ok, or the user declined. Report any issue verify lists.

5. Report

Give the log path, the counts of moved papers, fixed tags, and rewritten links, and the domains file path. Say that DeepPaperNote now follows the convention when it files a new paper, that the user can edit the domains file in Obsidian, and that "撤销上一次整理" / "undo the last reorganization" restores the library while the touched files stay unedited.

Related Skills

View on GitHub
GitHub Stars1.1k
CategoryEducation
Updated1mo ago
Forks78

Languages

Python

Trust signals

100/100

From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.

No cautions
organize-wiki — Universal Skill: Install & Safety Check | SkillAgent