content-ideate

name	content-ideate
description	Generate an article or tutorial concept from a source (tweet, article, or raw idea). Use when you have a piece of content or an idea and want to turn it into a lablab content concept — outputs a working title and brief ready to hand off to content-start.
compatibility	Designed for Claude Code. Requires WebSearch or Brave Search MCP for uniqueness check, WebFetch for URL sources, and python3 with requests installed.

Steps

1. Parse the source

Accept a tweet URL/text, article URL/text, or raw idea. Fetch URLs. Extract the core claim, tools mentioned, and implied audience.

2. Identify the angle

• Tutorial if there's something to build; Article if it's a trend, comparison, or opinion.
• Pick one reader level: Beginner / Intermediate / Advanced.

3. Deep uniqueness check

The goal is to find an angle no one has written yet. Run all of these:

1. site:lablab.ai [topic] — if covered, the angle must be meaningfully different
2. [topic] tutorial — broad web search, scan top 10 results across all platforms
3. [topic] site:medium.com OR site:dev.to OR site:hashnode.com OR site:towardsdatascience.com
4. [topic] site:youtube.com — video coverage signals a saturated angle; find what they missed
5. [specific tool/API combo] tutorial [current year] — check for recency gaps

After all searches, answer:

• What exact angles already exist? (list titles + platforms)
• What angle is missing entirely — a combination of tools, a use case, or a reader level no one has addressed?
• Is there a recency gap — old tutorials on a rapidly-changing tool?

Only proceed with an angle that passes: "I could not find this written anywhere." If everything is covered, reframe around the gap rather than retreading existing ground.

4. Generate titles

Produce 3 options — beginner-friendly, outcome-focused, comparison/contrast. Avoid: "A Guide to X", "Everything About X", "The Ultimate X Tutorial".

5. Present the brief

Type: [Article / Tutorial]
Recommended title: [title]
Alternative titles:
  - [option 2]
  - [option 3]

Hook: [one sentence — what makes this different from what already exists]
Target reader: [Beginner / Intermediate / Advanced]
Angle: [2-3 sentences — what the piece covers, what the reader walks away with]

Suggested file path: [blog/en/slug.mdx or tutorials/en/slug/article.mdx]

Wait for the user to confirm title and angle.

6. Post to Notion

Once confirmed, run:

python3 scripts/post_to_notion.py \
  --title "confirmed title" \
  --type "Article" \
  --hook "hook sentence" \
  --reader "Beginner" \
  --angle "angle summary" \
  --path "suggested/path.mdx"

Share the returned Notion URL with the user.

7. Create a design request for Damian

Immediately after posting to Notion, create a cover image request in the design database so Damian can start work in parallel.

Design database ID: 3485a26f46938097b6a9f4165323e8ce Assigned by (Stephen Kimoi): 2bed872b-594c-81d6-a2d4-0002487dc95b Default due date: 7 days from today

NOTION_TOKEN=$(grep NOTION_TOKEN /Users/steve/Lablab-AI/community-content/.agents/skills/.env | cut -d= -f2 | tr -d '"' | tr -d "'" | tr -d ' ')

curl -s -X POST "https://api.notion.com/v1/pages" \
  -H "Authorization: Bearer ${NOTION_TOKEN}" \
  -H "Notion-Version: 2022-06-28" \
  -H "Content-Type: application/json" \
  --data '{
    "parent": {"database_id": "3485a26f46938097b6a9f4165323e8ce"},
    "properties": {
      "Task name": {"title": [{"text": {"content": "<confirmed title>"}}]},
      "Status": {"status": {"name": "Ready for review"}},
      "Assigned by": {"people": [{"id": "2bed872b-594c-81d6-a2d4-0002487dc95b"}]},
      "Due date": {"date": {"start": "<YYYY-MM-DD>"}}
    },
    "children": [
      {"object":"block","type":"paragraph","paragraph":{"rich_text":[{"type":"text","text":{"content":"Type: <Article or Tutorial>"}}]}},
      {"object":"block","type":"paragraph","paragraph":{"rich_text":[{"type":"text","text":{"content":"Hook: <hook sentence>"}}]}},
      {"object":"block","type":"paragraph","paragraph":{"rich_text":[{"type":"text","text":{"content":"Angle: <2-3 sentence angle summary>"}}]}},
      {"object":"block","type":"paragraph","paragraph":{"rich_text":[{"type":"text","text":{"content":"Target reader: <Beginner / Intermediate / Advanced>"}}]}},
      {"object":"block","type":"paragraph","paragraph":{"rich_text":[{"type":"text","text":{"content":"Suggested file path: <path>"}}]}},
      {"object":"block","type":"paragraph","paragraph":{"rich_text":[{"type":"text","text":{"content":"Design notes: Cover image for lablab.ai article. Follow existing article cover style — clean, Lablab-branded, topic-relevant visual. Image will be uploaded to Cloudflare Images and added to the article frontmatter before publishing."}}]}}
    ]
  }'

Parse the url field from the response and share it with the user alongside the Notion content task URL.

If the API returns an error, report it clearly — do not silently skip this step.

Gotchas

• If the post_to_notion script exits with a 404, the database isn't shared with the integration yet. Ask the user to open the database in Notion → "..." → "Add connections" → "Claude Automation".
• --type must be exactly Article or Tutorial; --reader must be Beginner, Intermediate, or Advanced.
• Valid status for the design database is "Ready for review" — do not use "Not started" or "In Progress", they don't exist in that DB.

Hand off

Tell the user to run content-start with the confirmed title to begin the full pre-writing protocol.