agentleFS
iusztinpauliusztinpaul/designing-real-world-ai-agents-workshop/.agents/skills/research-and-write/SKILL.md

research-and-write

Skill512 starsChanged 7 months agoMITView on GitHub

End-to-end workflow: research a topic and then write a LinkedIn post about it. Use this skill whenever the user wants the full pipeline — from a topic idea to a finished LinkedIn post. Triggers on: 'research and write a post about', 'create a LinkedIn post about [topic]', 'I want to post about', 'write about [topic] for LinkedIn', or any request that implies both researching a subject and producing a LinkedIn post from it. This is the go-to skill when the user gives you a topic and expects a finished post.

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 2 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If the user only gives a topic, ask for the guideline details (angle, audience, key points, tone) or suggest a default based on the topic.

    line 16
  2. Very specific:

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

    line 20
  3. Distinctive:

    Create the directory if it doesn't exist.

    line 24
  4. Very specific:

    Create guideline.md in the working directory:

    line 26
  5. Very specific:

    Use outputs/{slug}/ as the working_dir for all tool calls.

    line 56
  6. Very specific:

    Read the WORKFLOW_INSTRUCTIONS from src/writing/routers/prompts.py and follow those steps exactly, using the linkedin-writer MCP tools.

    line 62

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Research and Write

End-to-end workflow: research a topic, then write a LinkedIn post from it. Chains the deep-research and linkedin-writer MCP servers.

Input Preparation

Gather from the user:

  1. Topic — what to research
  2. Guideline — how the post should be written (becomes guideline.md)

If the user only gives a topic, ask for the guideline details (angle, audience, key points, tone) or suggest a default based on the topic.

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.

Create guideline.md in the working directory:

# LinkedIn Post Guideline

## Topic
[Core topic]

## Angle
[Perspective]

## Target Audience
[Who reads this]

## Key Points to Cover
[3-5 bullets]

## Tone
[How it should sound]

Execution

Phase 1: Research

Load the research_workflow MCP prompt from the deep-research server and follow the workflow instructions using the available tools:

  • deep_research — for web research queries
  • analyze_youtube_video — for any YouTube URLs the user provides
  • compile_research — to produce the final research.md

Use outputs/{slug}/ as the working_dir for all tool calls. This produces research.md.

Tell the user when research is complete.

Phase 2: Write

Read the WORKFLOW_INSTRUCTIONS from src/writing/routers/prompts.py and follow those steps exactly, using the linkedin-writer MCP tools. The working directory outputs/{slug}/ already has guideline.md and research.md from Phase 1.

The generate_post tool internally runs 4 evaluator-optimizer iterations (review + edit cycles) to refine the post before producing the final version.

After Completion

Present the final outputs/{slug}/post.md and outputs/{slug}/post_image.png to the user. Offer to edit with feedback.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.