Agent skill

Eva Tweet

by vvedantb in vvedantb/eva

Write launch/announcement tweets in Vedant's style — terse dev-tool shipping posts built from a one-line hook, an optional positioning line, an arrow bullet list of capabilities, and a punchy close.

MITAuto-check passedWriting & Content

Install Eva Tweet

skills CLI
$ npx skills add vvedantb/eva --skill eva-tweet -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install vvedantb/eva eva-tweet --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/vvedantb/eva.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.claude/skills/eva-tweet .claude/skills/eva-tweet && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
eva-tweet
GitHub stars
101
Token cost
~1.6k tokens
SKILL.md length
769 words
Files
1
Skills in repo
25
Repo updated
First seen
Licence
MIT

At a glance

Write launch/announcement tweets in Vedant's style — terse dev-tool shipping posts built from a one-line hook, an optional positioning line, an arrow bullet list of capabilities, and a punchy close.

  • Editing a tweet
  • SKILL.md covers The shape, Block 1 — the hook, Block 2 — positioning and Block 3 — the arrow list, plus 6 more sections
  • Calls just
  • Launch/changelog post announcing a feature

What it does

Eva Tweet is an agent skill from vvedantb/eva. Write launch/announcement tweets in Vedant's style — terse dev-tool shipping posts built from a one-line hook, an optional positioning line, an arrow bullet list of capabilities, and a punchy close. Use whenever drafting or editing a tweet, X post, thread, or launch/changelog post announcing a feature, release, version bump, or experiment — even if he doesn't say "in my voice".

Its SKILL.md is about 1.6k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Writing & Content, covering Social media posts. It works with X (Twitter). The repository describes itself as: Orchestrate sandboxed agents that run in the cloud while you work. The licence is MIT.

When your agent uses it

  • Editing a tweet
  • Launch/changelog post announcing a feature
  • Experiment — even if he doesnt say in my voice

Example prompts

  • “t say”
  • “/eva-tweet”

What it can do on your machine

Read from SKILL.md and the folder at commit 1c4532e. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • just

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Eva Tweet loads about 1.6k tokens when it runs. Until then it costs about 98 tokens; SKILL.md has 769 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~98
When it runs · the whole SKILL.md, loaded when a task matches
~1.6k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from vvedantb/eva at commit 1c4532e, republished under its MIT licence (© vvedantb). 769 words, ~1,559 tokens.

Download SKILL.mdSave it as .claude/skills/eva-tweet/SKILL.md (or your agent's skills folder).
name
eva-tweet
description
Write launch/announcement tweets in Vedant's style — terse dev-tool shipping posts built from a one-line hook, an optional positioning line, an arrow bullet list of capabilities, and a punchy close. Use whenever drafting or editing a tweet, X post, thread, or launch/changelog post announcing a feature, release, version bump, or experiment — even if he doesn't say "in my voice".

Vedant's tweet voice

These are dev-tool shipping posts. The reader is another engineer scrolling fast. The tweet's job is to state what now exists and what it can do — nothing else. No selling, no story, no thread of context.

The whole style is one idea: capability lists do the persuading. You never write that something is fast or powerful; you write 83ns update dispatch and let the reader conclude it.

This is a written voice and it is not the same as [[vedant-voice]], which is spoken, warm, and story-shaped (before→now→why). Do not import that structure here. Tweets are flat, clipped, and declarative.

The shape

Almost every post is the same four blocks, separated by blank lines. Blocks 2 and 4 are optional.

1. Hook          — what now exists, one line
2. Positioning   — what it is, one fragment (no verb needed)
3. Arrow list    — 3–5 capabilities, each prefixed →
4. Close         — the goal, the payoff, or a fragment triad

Worked example:

Introducing <terminal />

A built-in component for Native SDK

→ Real shell on a PTY
→ Powered by libghostty-vt
→ Selection, scrollback + truecolor
→ Record sessions, replay them offline

Give your software factory or agent orchestrator a terminal

Short posts drop straight to blocks 1–2, or 1 alone:

portless v0.15.3 is out

HTTPS from WSL now just works in Windows browsers

Block 1 — the hook

One line. It names the thing and the fact that it exists now. Pick the opener that matches what actually happened:

  • Introducing X — a genuinely new product or surface
  • New experiment: X — early, unproven, explicitly a punt
  • Now available: X — a feature landing in something that already shipped
  • X now <verbs> — an existing thing gained a capability ("Native SDK now renders CJK")
  • You can now <do X> — framed from the user's side
  • X v0.15.3 is out — a version bump

Never We're excited to announce. Never 🚀. Never a question as the hook.

Block 2 — positioning

One fragment, no full sentence needed, no period required. It answers "what is this" for someone who's never heard of it.

  • A TypeScript-to-native compiler
  • A built-in component for Native SDK
  • agent-browser is now a browser agent.

If the hook already made it obvious, cut this block.

Block 3 — the arrow list

The core move. → then a terse capability. Rules:

  • 3–5 items. Six is too many, two isn't a list.
  • No trailing punctuation. Ever.
  • Fragments, not sentences. → Type checked by tsc, not → It is type checked by tsc.
  • Sentence case, except where a lowercase project name or all-lowercase enumeration is natural: → inputs, trees, charts, dialogs.
  • Mix the register freely — a list can hold a technical spec, a benefit, and a get-out clause side by side: → 83ns update dispatch / → Real TS: Classes, generics, try/catch / → No Zig required / eject to Zig anytime.
  • Lead with what's most surprising, not what's most fundamental.

Some items are literal commands or invocations rather than prose — that's fine and common: → One-shot: agent-browser chat "open google, search for dogs".

Block 4 — the close

One line, optional. It's either:

  • the goal, stated flatly — Goal: compile existing TS/JS with no code changes
  • the payoff, imperative — Give your software factory or agent orchestrator a terminal
  • a fragment triad — three clipped fragments, usually two negations then the positive:
    No webview. No DOM.
    Just markup + TypeScript.
  • a benefit line with numbers — For agents: half the turns, half the cost, half the LOC

Never a call to action. No "check it out", no "link below", no "let me know what you think".

Show full SKILL.md (287 more words)Show less

Code and commands

CLI goes bare on its own line — no backticks, no code fences, no indentation. Twitter renders them as literal characters, so they add noise.

agent-browser a11y <url>

Multi-line snippets can carry a # comment to mark the human step:

agent-browser network har start
# agent browses, clicks around
agent-browser network har stop

Register

  • Terse. Declarative. Present tense. Every line earns its place or gets cut.
  • Concrete specifics always — 40+ built-in components, 83ns, WCAG, axe-core, v0.15.3, libghostty-vt. Named dependencies are credibility.
  • "just" as a minimizer carries over from his spoken voice: HTTPS from WSL now just works, Just markup + TypeScript.
  • Lowercase project names stay lowercase — agent-browser, portless, json-render, eva. Never title-case them at the start of a line.
  • Blank line between every block. The whitespace is the layout.

What to avoid

  • No hashtags. No @-mentions. No links. None of the reference posts use them.
  • No exclamation marks. Emoji almost never — at most one, only where something is genuinely wry.
  • No adjective-selling — "powerful", "blazing fast", "seamless", "game-changing", "revolutionary". State the spec instead.
  • No hype openers — "We're excited to", "Big news", "Thrilled to share".
  • No threads by default. One post, one shipped thing. Only thread if the user asks.
  • No before→now story. That's the spoken voice. Here you state the after and stop.

Length

Under 280 characters, and usually well under. Aim for the shortest version that still names every capability worth naming. If it doesn't fit, cut an arrow item — never compress by removing whitespace.

How to apply

Given a feature or release, ask three questions: what now exists (hook), what is it (positioning), what can it do (3–5 arrows). Write those, then decide whether a close adds anything — if not, end on the list. Strip every adjective that isn't load-bearing, then check each arrow item for a trailing period and delete it.

© vvedantb, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .claude/skills/eva-tweet of vvedantb/eva.

Open the folder on GitHubat commit 1c4532e

Compare with similar skills

Eva Tweet next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Eva Tweet compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Eva Tweet this skillvvedantb/eva101—~1.6kAutomated safety check: PassMIT
Socialcoreyhaines31/marketingskills54k4 repos~4.5kAutomated safety check: PassMIT
Outlier Post FinderScrapeCreators/social-media-research-skills3.4k—~1.5kAutomated safety check: NotesMIT
X Algorithm Post Writingcarson2222/skills113—~3.8kAutomated safety check: PassApache-2.0
X Researchtrailofbits/skills-curated513—~1.9kAutomated safety check: NotesCC-BY-SA-4.0
X Founder OperationsLeoYeAI/openclaw-master-skills2.2k—~4.6kAutomated safety check: PassMIT

Similar skills

  • Social

    coreyhaines31/marketingskills

    When the user wants help creating, scheduling, or optimizing social media content for LinkedIn, Twitter/X, Instagram, TikTok, or Facebook, or wants to do social listening and engagement triage.

    54k GitHub starsUsed in 4 repos~4.5k tokens
    Writing & ContentAuto-check passed
  • Outlier Post Finder

    ScrapeCreators/social-media-research-skills

    A skill your agent uses when the user wants to find posts, videos, reels, shorts, tweets, or social content that overperformed versus a creator, brand, or competitor baseline.

    3.4k GitHub stars~1.5k tokensUpdated 1 mo ago
    Writing & ContentAuto-check: notes
  • X Algorithm Post Writing

    carson2222/skills

    Writes and reviews X posts, threads and replies using what the open-sourced For You ranking system rewards, and explains why a post may have underperformed.

    113 GitHub stars~3.8k tokensUpdated 3 mo ago
    Writing & ContentAuto-check passed
  • X Research

    trailofbits/skills-curated

    Official

    Searches X/Twitter for real-time perspectives, dev discussions, product feedback, breaking news, and expert opinions using the X API v2.

    513 GitHub stars~1.9k tokensUpdated 2 mo ago
    Writing & ContentAuto-check: notes
  • X Founder Operations

    LeoYeAI/openclaw-master-skills

    Systematic X (Twitter) operations skill for founders, indie developers, and tech professionals.

    2.2k GitHub stars~4.6k tokensUpdated 2 mo ago
    Writing & ContentAuto-check passed
  • Launch Distribution

    amplitude/builder-skills

    Plan and execute multi-channel distribution for product launches across X/Twitter, Reddit, LinkedIn, Hacker News, and other platforms.

    159 GitHub stars~2.8k tokensUpdated 2 mo ago
    Writing & ContentAuto-check passed

More from vvedantb/eva

All 25 skills in this repo
  • Brag

    vvedantb/eva

    Turn the current project website into a short, polished, shareable launch video using Hyperframes.

    101 GitHub starsUsed in 2 repos~2.3k tokens
    Auto-check passed
  • Eva Feature Demo

    vvedantb/eva

    Record a real agent-browser screencast of one eva feature being used end to end, convert it to an X-ready mp4, and write a tweet for it.

    101 GitHub stars~3.3k tokensUpdated today
    Auto-check passed
  • Animate

    vvedantb/eva

    Build an animation from scratch, making the decisions in the order that determines whether it feels right — should it animate at all, what purpose, which tool, which properties, which curve and…

    101 GitHub starsUsed in 6 repos~2.9k tokens
    Auto-check passed
  • Design and build Convex components with clear boundaries, isolated state, and app-facing wrappers.

    101 GitHub stars~3.3k tokensUpdated today
    Auto-check passed
  • Grab a single clean HD screenshot of a new eva feature from the real running app (Playwright at deviceScaleFactor 2, 1280 layout captured crisp at 2560×1440, dev overlays hidden) and write a tweet…

    101 GitHub stars~3.2k tokensUpdated today
    Auto-check: notes
  • Code Structure

    vvedantb/eva

    A skill your agent uses when multiple workflows duplicate the same operational logic, when deciding what belongs in actions vs shared services, or when refactoring repeated operational blocks across…

    101 GitHub starsUsed in 2 repos~1.1k tokens
    Auto-check passed

Works with

Questions about Eva Tweet

What does Eva Tweet do?

Write launch/announcement tweets in Vedant's style — terse dev-tool shipping posts built from a one-line hook, an optional positioning line, an arrow bullet list of capabilities, and a punchy close. Eva Tweet is an agent skill from vvedantb/eva. Write launch/announcement tweets in Vedant's style — terse dev-tool shipping posts built from a one-line hook, an optional positioning line, an arrow bullet list of capabilities, and a punchy close.

When should I use Eva Tweet?

Eva Tweet fits situations like: editing a tweet; launch/changelog post announcing a feature; experiment — even if he doesnt say in my voice.

How do I install Eva Tweet in Claude Code?

Run `npx skills add vvedantb/eva --skill eva-tweet -a claude-code`. Or copy the skill folder (.claude/skills/eva-tweet in vvedantb/eva) into .claude/skills/eva-tweet in your project. Claude Code loads it when a task matches its description.

How do I install Eva Tweet in Codex?

Run `npx skills add vvedantb/eva --skill eva-tweet -a codex`. Or copy the skill folder (.claude/skills/eva-tweet in vvedantb/eva) into .agents/skills/eva-tweet in your project. Codex loads it when a task matches its description.

Can I use Eva Tweet in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add vvedantb/eva --skill eva-tweet -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/eva-tweet, .gemini/skills/eva-tweet, .github/skills/eva-tweet and .opencode/skills/eva-tweet in your project.

What does Eva Tweet need to run?

Going by SKILL.md and its folder, Eva Tweet needs the command-line tools its instructions call (just).

Does Eva Tweet access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Eva Tweet safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Eva Tweet use?

Eva Tweet is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Eva Tweet use?

About 1.6k tokens (SKILL.md is roughly 6.2k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Eva Tweet?

Skills that share tags, products or a category with Eva Tweet: Social (coreyhaines31/marketingskills, 54k stars), Outlier Post Finder (ScrapeCreators/social-media-research-skills, 3.4k stars), X Algorithm Post Writing (carson2222/skills, 113 stars) and X Research (trailofbits/skills-curated, 513 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Eva Tweet?

vvedantb (a GitHub user) maintains it in vvedantb/eva, which has 101 GitHub stars. The repository holds 25 skills in this directory. The repository was last updated on October 10, 2026.

Source: vvedantb/eva on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.