User types: /ingest <url|text>
Categories: article | doc | paper | note
Step 1 — FETCH
Fetch the URL, strip nav/ads/footers/JS/cookie banners
Convert to clean markdown, preserve headings, paragraphs, code blocks
Download inline images → raw/assets/
Build frontmatter:
title:
url:
author: <extracted or "unknown">
published: <extracted date or "unknown">
fetched:
category: <article|doc|paper|note>
status: raw
Save to the correct folder based on category:
article → raw/articles/.md
doc → raw/docs/.md
paper → raw/papers/.md
note → raw/notes/.md
Category-Specific Notes
article: general tech articles, blog posts, tutorials
doc: reference/official documentation — preserve structure more literally, less summary
paper: academic — extract abstract, methodology, findings as distinct sections
note: pasted text or local file path instead of URL; minimal processing
Step 2 — LOG THE FETCH
Append to wiki/log.md:
[YYYY-MM-DD] ingest |
Fetched from . Saved to raw/s/.md. Status: raw.
Step 3 — AUTO COMPILE
Immediately execute /compile on the saved raw file
Load and follow: .claude/skills/compile/SKILL.md
Do NOT wait for user confirmation before compiling
Notes
raw/ is READ ONLY after save — /compile reads it, never modifies content
Only status: field in frontmatter changes (raw → compiled) during /compile
User can manually call /compile later: /compile raw/s/.md
Batch compile an entire folder: /compile raw/articles/
1---2name: ingest3description: Skill: /ingest4---5# Skill: /ingest67## Trigger8User types: /ingest <category> <url|text>9Categories: article | doc | paper | note1011---1213## Step 1 — FETCH14- Fetch the URL, strip nav/ads/footers/JS/cookie banners15- Convert to clean markdown, preserve headings, paragraphs, code blocks16- Download inline images → raw/assets/<filename>17- Build frontmatter:18 title: <extracted page title>19 url: <source url>20 author: <extracted or "unknown">21 published: <extracted date or "unknown">22 fetched: <today YYYY-MM-DD>23 category: <article|doc|paper|note>24 status: raw25- Save to the correct folder based on category:26 article → raw/articles/<slugified-title>.md27 doc → raw/docs/<slugified-title>.md28 paper → raw/papers/<slugified-title>.md29 note → raw/notes/<slugified-title>.md3031## Category-Specific Notes32- article: general tech articles, blog posts, tutorials33- doc: reference/official documentation — preserve structure more literally, less summary34- paper: academic — extract abstract, methodology, findings as distinct sections35- note: pasted text or local file path instead of URL; minimal processing3637## Step 2 — LOG THE FETCH38Append to wiki/log.md:39 ## [YYYY-MM-DD] ingest | <title>40 Fetched from <url>. Saved to raw/<category>s/<slug>.md. Status: raw.4142## Step 3 — AUTO COMPILE43- Immediately execute /compile on the saved raw file44- Load and follow: .claude/skills/compile/SKILL.md45- Do NOT wait for user confirmation before compiling4647## Notes48- raw/ is READ ONLY after save — /compile reads it, never modifies content49- Only status: field in frontmatter changes (raw → compiled) during /compile50- User can manually call /compile later: /compile raw/<category>s/<slug>.md51- Batch compile an entire folder: /compile raw/articles/
Run npx skillmds@latest add lokesh-go/ingest in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.
Skill: /ingest It is listed under Coding & Dev Tools on SkillMD.
This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.
This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.
Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.
lokesh-go (@lokesh-go) published this skill. Their other Agent Skills are listed on their SkillMD profile.
lokesh-go