Research Content

Conduct deep research to inform content creation. This skill gathers context, finds relevant information, analyzes trends, and produces structured briefs that feed directly into content writing.

Step 1: Understand the research request

Ask the user:

1. Topic - what are we researching?
2. Platform - Twitter, GitHub, blog, LinkedIn, other?
3. Goal - announce a project, share insight, grow audience, educate, sell?
4. Audience - developers, crypto community, founders, general tech, investors?
5. Sector - dev, crypto, AI/ML, startup, open source, design?
6. Existing content - any references, repos, or content to build on?
7. Competition - want to see what similar projects/people are doing?

Skip questions you can confidently infer. If someone says "research my new Rust CLI tool for dev Twitter," you already know the platform, sector, and audience.

Step 2: Conduct the research

Project research

If there's a specific project, repo, or product:

• Analyze the repo (README, code structure, tech stack, stars, activity)
• Read documentation and changelogs
• Identify the unique value prop - what makes this different?
• Find the "aha moment" that would hook the target audience
• Note installation complexity (one-liner? multi-step?)

Audience research

• What content performs well in this community?
• What vocabulary does the audience use? (see sector-profiles.md)
• What are they frustrated about? Excited about?
• Who are the key voices? What do they post about?
• What content gaps exist - what's NOT being said that should be?

Competitive research

• How do similar projects/products communicate?
• What README formats are competitors using?
• What tweet styles perform best for similar announcements?
• What can we do differently to stand out?

Trend research

Use WebSearch to find:

• Current trending topics in the relevant sector
• Recent news or developments the content could hook into
• Upcoming events, launches, or milestones to time against
• Hashtags and keywords with momentum

Platform-specific research

For Twitter/X: Read twitter-algorithm.md for the current algorithm mechanics. Key points:

• Replies are weighted 27x a like, reply-to-reply 150x
• Bookmarks and dwell time are 20x a like
• External links get 30-50% reach reduction - use link-in-reply strategy
• First 30-60 minutes are the critical engagement window
• 2-5 quality tweets/day, spaced 30-60 min apart
• Premium subscribers get roughly 10x more reach per post

Research for the specific content:

• High-performing tweet formats for this sector
• Key accounts to engage with (for reply strategy)
• Optimal thread length for this topic
• Whether to use Premium long-form vs thread vs single tweet

For GitHub:

• Similar projects and their README approaches
• What makes top repos in this space successful (stars, forks, contributor count)
• Badge and shield recommendations for credibility
• Documentation patterns that work in this niche
• Community engagement strategies (discussions, issues, contributing guides)

For LinkedIn:

• Professional framing of the same content
• Post length sweet spots (1300 chars for feed posts)
• Hook patterns that work on LinkedIn (first 2 lines visible)
• Carousel/document post potential

For Blog/Newsletter:

• SEO keyword research
• Content format (tutorial, opinion, case study, comparison)
• Linking strategy
• Distribution plan

Step 3: Produce the research brief

Output a structured brief the user can act on immediately:

Key findings

5-10 bullet points of actionable insights. Not vague observations but specific things the user can use.

Content angles

3-5 different approaches to the topic. Each angle should have a one-sentence pitch and a note on expected engagement type (likes vs replies vs bookmarks vs shares).

Audience insights

Who cares about this and why. Specific pain points, desires, knowledge level. What vocabulary they use.

Competitive landscape

What others are saying. Gaps and opportunities. What's overplayed and what's underserved.

Recommended format

Best content format for this topic + platform combo. Thread vs single tweet, short README vs detailed, etc. With reasoning.

Key terms and phrases

Vocabulary the target audience uses naturally. Community-specific slang, jargon, and references.

Risks and things to avoid

What NOT to say or do. Sensitive topics, overplayed angles, common mistakes in this niche.

Sources

All URLs and references consulted.

Step 4: Offer next steps

After delivering the brief, suggest:

• "Want me to draft content based on this research?" (triggers write-content skill)
• "Want me to dig deeper into any angle?"
• "Want me to research competitor accounts for reply strategy?"

The research brief should be detailed enough that the write-content skill can produce high-quality content from it without additional research.