write-post

Write LinkedIn Post

Generate a LinkedIn post using the linkedin-writer MCP server.

Working Directory

All output goes into outputs/{slug}/ relative to the project root. Derive the slug from:

• The dataset seed/guideline filename if the user references one (e.g., my-topic_seed.md → my-topic)
• Otherwise, slugify the topic (lowercase, hyphens, no special chars, max 60 chars)

Create the directory if it doesn't exist.

Input Preparation

The working directory needs guideline.md and research.md.

If the user provides raw text for the guideline, create guideline.md in the working directory:

# LinkedIn Post Guideline

## Topic
[What the post is about]

## Angle
[What perspective or approach to take]

## Target Audience
[Who this post is for]

## Key Points to Cover
[3-5 bullet points]

## Tone
[How it should sound]

If research.md is in a different location, copy it into the working directory.

Execution

Read the WORKFLOW_INSTRUCTIONS from src/writing/routers/prompts.py and follow those steps exactly, using the linkedin-writer MCP tools. Pass outputs/{slug}/ as the working directory path to each tool.

After Completion

Present the final outputs/{slug}/post.md content to the user. If an image was generated, mention outputs/{slug}/post_image.png.