brainstorm-linkedin

Brainstorm LinkedIn Posts from Source Material

Generate LinkedIn post ideas based on external context provided by the user.

Process

1. Read ALL published posts (MANDATORY - to avoid topic/angle overlap): ```bash ./scripts/print-published.sh linkedin-post ``` This prints all published posts with full content in one call. Note:
2. Core insights already covered
3. Data points already used
4. Angles already explored

Do not suggest ideas that repeat existing coverage.

1. Read the style guide:
2. guidelines/linkedin.md
3. references/professional-profile.md

1. Process the provided context (user will provide one or more of):
2. Files (PDFs, documents, research papers)
3. URLs (articles, blog posts, announcements)
4. Raw text or ideas

1. Identify 2-4 promising angles by considering:
2. What's the unique insight for a professional audience?
3. How does this connect to the author's expertise?
4. What's the hook that works in the first 210 characters?
5. Is there a personal angle or company connection?

1. Present ideas with for each:
2. Proposed title
3. Core insight (1 sentence)
4. Hook approach (personal anecdote, company experience, surprising outcome, or news)
5. Why it fits the author's voice

1. Ask user to choose:
2. Which idea(s) to develop
3. Whether to create as idea (01-ideas) or draft (02-drafts)

Evaluation Criteria

Strong LinkedIn post ideas have:

• A concrete hook in the first 210 characters
• A clear insight that provides value
• Connection to the author's expertise areas
• Room for a personal or company angle
• Appropriate scope for the target word count

Creating Files

After user selection, get timestamp:

date -u +"%Y%m%d-%H%M%S"  # For slug
date -u +"%Y-%m-%dT%H:%M:%S.000Z"  # For created/lastUpdated

Create file at: content/posts/linkedin-post/{slug}-{slugified-title}.yaml