wiki-ingest
Processes and ingests new documents, PDFs, links/URLs, HTML, or Markdown files into the Wiki.
Skill by not-a-feature
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/not-a-feature-wiki-ingest
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: documents, frontend
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/wiki-ingest
What's inside
- Skill name
wiki-ingest- Description
- Processes and ingests new documents, PDFs, links/URLs, HTML, or Markdown files into the Wiki. Use when the user drops a new file or refers to a new source to be added to the knowledge base.
- License
- LGPL-2.1
- Instructions
- 87 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Wiki Ingest Skill Two workflows, depending on the input. --- ## Branch A: Literature & Document Ingestion *A raw file or a URL to a paper/article (PDF, HTML, Markdown, text), or a request to ingest a specific reference.* > [!TIP] > For automating ingestion (especially of PDFs), you can run the library script locally in a two-step process: > 1. Run the script with `--extract-only` to download the file (if it's a URL) and extract its text: > `python Scripts/library.py <file_path_or_url> --extract-only` > 2. The script outputs the local path of the file (under `LOCAL_PATH: ...`) and the extracted text. Read the text, perform metadata extraction manually as the LLM (following the schema in A3), and save the structured metadata to a temporary JSON file inside the conversation's scratch directory: `<appDataDir>\brain\<conversation-id>\scratch/metadata.json`. > 3. Complete the ingestion by running the script again with `--metadata`, `--no-confirm`, and `--delete-original`: > `python Scripts/library.py <local_path_returned> --metadata <path_to_metadata_json> --no-confirm --delete-original` > This automatically handles file renaming/copying, markdown note creation, and index updating. Proceed to Branch B if needed. ### A1: Locate Source - URL (starts with `http`): download to `Wiki/Library/tmp.<ext>`. Use `.pdf` for arbitrary/binary URLs; preserve `.html`/`.md`/`.txt` if present in the path. - Local path: locate it. ### A2: Extract Text Read Markdown/text directly. For PDF/HTML, parse text cleanly, ignoring binary/markup wrapping. You may use the command line tool "markitdown" for this. ### A3: Extract Metadata 1. **title:** APA-style Title Case. 2. **filename:** from title + year, descriptive, under 40 chars (e.g. `Attention_Is_All_You_Need_2017`). Replace spaces, hyphens, slashes, backslashes, dots, question marks, curly braces, and `'s` with single underscores; strip double/leading/trailing underscores. 3. **tag:** one category from `default_tags` in the plugin's `tags.json`. 4. **tag_list:** up to 5 PascalCase tags for indexing, excluding any in `negative_tags` (`tags.json`). 5. **abstract:** exact abstract from the paper; replace unicode with ASCII/LaTeX equivalents; inline math as `$...$`. 6. **topic1-4:** one short sentence each, starting with an action verb - main method, second method, discussion, results (e.g. "Presents..."). ### A4: Confirm with User - Check whether the document is relevant to existing projects, collaborators, or pages - i.e. whether Branch B synthesis applies. - Present the metadata (Title, Filename, Tag, Tag list, Abstract, Topics) then ask: **"Is this correct?"** ### A5: Save, Index, Synthesize (once confirmed) 1. **Copy source** to `Wiki/Library/<filename>.<ext>`; delete the temp download if used. 2. **Branch B (if confirmed):** run steps B2 and B3 using the paper's contents. 3. **Markdown note (PDFs only)** at `Wiki/Library/<filename>.md`: ```markdown >[!abstract]- > <abstract> - :luc_file_text: [[<filename>.pdf]] - <topic1> - <topic2> - <topic3> - <topic4> #<tag> #<tag_list_item1> #<tag_list_item2> ... ------------- ``` 4. **Index** — prepend to the top of `Wiki/_Index.md`: ```markdown ### <title> >[!abstract]- > <abstract> - :luc_file_text: [[<filename>.pdf]] ❌ - :luc_file: [[<filename>.md|Notes]] - <topic1> - <topic2> - <topic3> - <topic4> ``` *(Non-PDF source: omit the `- :luc_file:` line.)* --- ## Branch B: Research Notes & Project Ingestion *Raw research notes, discussion questions, project updates, or meeting minutes.* ### B1: Synthesize Read the input to identify the core concept, project updates, methods, or open questions. Clarify anything ambiguous with the user. ### B2: Create or Update Pages Draft or modify the relevant page in `Wiki/Projects/` or `Wiki/`. Merge rather than overwrite — note whether new input refines, expands, or contradicts prior claims. Add Obsidian links (`[[Link]]`) connecting the concept to existing projects, collaborators, and methods. ### B3: Update Navigation - Prepend to `Wiki/log.md`: `## [YYYY-MM-DD HH:MM] ingest | <Title>`. - If new pages were created, add their links to `Wiki/wiki.md` under the right category. ## Reminders - Only touch `Wiki/_Index.md` and `Library` via the steps above; do not otherwise hand-edit them. - Preserve existing cross-references when editing a page. - Tone: dry, objective, scientific, concise.
From not-a-feature/wiki-skill, LGPL-2.1.
More from not-a-feature/wiki-skill
- 0.474score
wiki-person
not-a-feature
Creates or updates a person profile in the Wiki's Persons directory following the standardized framework.
SkillCFreshness: CClaude Code +9 - 0.471score
wiki-report
not-a-feature
Generates a summary report of progress, collaborator updates, and open questions over a given timeframe (defaulting to 8 days) based on the PhD Wiki logs…
SkillCFreshness: CClaude Code +9 - 0.469score
wiki-lint
not-a-feature
Performs health checks and janitorial tasks on the Wiki.
SkillCFreshness: CClaude Code +9 - 0.469score
wiki-query
not-a-feature
Explores and synthesizes information from the Wiki based on a user's question.
SkillCFreshness: CClaude Code +9 - 0.469score
wiki-log
not-a-feature
Logs progress and updates regarding one project into the Wiki.
SkillCFreshness: CClaude Code +9 - 0.467score
wiki-daily
not-a-feature
Generates a daily brief summarizing active projects, open tasks, todos, open questions, and reminders.
SkillCFreshness: CClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Jul 8, 2026 (3 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- LGPL-2.1
- Repository
- github.com/not-a-feature/wiki-skill
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from not-a-feature
- 0.474score
wiki-person
not-a-feature
Creates or updates a person profile in the Wiki's Persons directory following the standardized framework.
SkillCFreshness: CClaude Code +9 - 0.471score
wiki-report
not-a-feature
Generates a summary report of progress, collaborator updates, and open questions over a given timeframe (defaulting to 8 days) based on the PhD Wiki logs…
SkillCFreshness: CClaude Code +9 - 0.469score
wiki-lint
not-a-feature
Performs health checks and janitorial tasks on the Wiki.
SkillCFreshness: CClaude Code +9 - 0.469score
wiki-query
not-a-feature
Explores and synthesizes information from the Wiki based on a user's question.
SkillCFreshness: CClaude Code +9 - 0.469score
wiki-log
not-a-feature
Logs progress and updates regarding one project into the Wiki.
SkillCFreshness: CClaude Code +9 - 0.467score
wiki-daily
not-a-feature
Generates a daily brief summarizing active projects, open tasks, todos, open questions, and reminders.
SkillCFreshness: CClaude Code +9
Cite this page
wiki-ingest skill. Agent Mart. https://agentmart.dev/skills/not-a-feature-wiki-ingest (last updated 2026-10-24)