Index of every page on this site: llms.txt
All skills
Run it yourself

Find tweet ideas, plan launch posts, and draft or polish tweets with code cards, stat cards, or screenshot wraps. Use for content strategy, what to post, when to post, or an X post.

  • 6 files
  • 42.7 KB
  • MIT
  • Updated last month
  • GitHub
Use this Skill: https://skilld.dev/gh/harlan-zw/harlan-agent-kit/tweet

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

referencesplaybook.md

≈783 tokens on demand. Your agent reads this file only when SKILL.md points to it.

Tweet content planning

Use these formats as starting points. Match the user's voice and audience. They are editorial suggestions, not verified engagement benchmarks.

Principles

  • Lead with a reader benefit, a concrete problem, or a useful observation.
  • Include a visual recommendation when it adds evidence or explains the topic.
  • Keep sentences short and separate distinct ideas.
  • Preserve the user's voice. Avoid corporate launch language.
  • Keep a single post unless the user wants a thread.
  • Suggest at most two relevant accounts. Verify handles before recommending them.
  • Never invent experience or engagement figures.
  • Apply the timeline test: would a reader care without already following the author?

Content Formats That Work

Project Launch

[One-line hook: what problem this solves]

[Brief description, 1-2 sentences max]

[Link or "link below"]
[Optional: screenshot/demo that explains the point]

Version Update / Changelog

Happy Friday! [Project] v[X.Y] just dropped:

- [Feature 1: framed as benefit]
- [Feature 2: framed as benefit]
- [Feature 3: framed as benefit]

[Link]
[Optional: screenshot of headline feature]

Technical Insight

[Non-obvious observation or lesson learned]

[1-2 sentences of context]

[Optional: code snippet screenshot]

Build-in-Public

[What you're working on + why]

[One interesting detail or challenge]

[Optional: screenshot of current state, even if rough]

Curated Share

[Your take on why this matters]

[Quote tweet or link to the thing]

Hooks That Stop the Scroll

Strong openers (adapt to your voice):

  • "Free tool that [does specific thing]"
  • "[Surprising stat or fact about a common problem]"
  • "TIL [non-obvious technical insight]"
  • "Watch: [brief description of what the demo shows]"
  • "[X] sites use [Y] wrong. Here's what they miss:"

Weak openers to avoid:

  • "Excited to announce..."
  • "After months of hard work..."
  • "Thread on [topic] 🧵"
  • "I just released..."
  • An opening about the author that gives the reader no useful context

Content mix

If the user wants a recurring plan, offer this mix as an optional starting point:

  • 50% project updates and launches.
  • 25% technical insights and tips.
  • 15% build-in-public stories.
  • 10% community engagement.

Adapt the mix to the user's goals and available material. Do not manufacture weak posts to fill a quota.

Cadence and timing

Ask for cadence only when it changes the requested plan. Use a manageable schedule based on the user's available time. Batch drafts when useful. Posting and scheduling require separate approval.

Prefer the user's audience analytics for timing. If analytics are unavailable, suggest a trial window and label it as a hypothesis. Use the audience's timezone and convert it for the requested date. Account for daylight saving. Never present a fixed AEST conversion as valid all year.

Follow-up ideas

Suggest a follow-up when it adds a new example, result, lesson, or answer. Avoid repeating the launch headline in another form. Never reply to another account without explicit approval.

Source: SKILL.md on GitHub

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

No third-party reports yet.

Signed by skilld at 2a530d0. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 2 days ago.

Activeupdated last month
argument-hint
[draft, topic, release notes, screenshot, or ideas]

README badge

README badge for harlan-zw/harlan-agent-kit/tweet