Guide · Agent skill
How to use the X / Twitter Automation agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The X / Twitter Automation skill (hive.x-automation) is a playbook that teaches an AI agent how to operate the X website through a browser: composing and posting, attaching images, replying to posts, and checking that the Post button is enabled before it clicks. It is written for developers, marketers and community managers who already run an agent with browser tools. After this guide you will know what the skill covers, how to install it, how to trigger it, and what it needs and risks before you run it.
X / Twitter Automationaden-hive/hive · Apache-2.0 Download the skill Quick start
- 1Download and unzip. Download the skill ZIP from this page and unzip it to get SKILL.md.
- 2Place the skill folder. Put it in ~/.claude/skills/ for personal use, in .claude/skills/ for one project, or upload the ZIP under Settings → Capabilities → Skills.
- 3Install browser-automation. Make sure the hive.browser-automation skill is also installed, since this skill requires it.
- 4Log in to X in the browser. Start your browser tool session and sign in to X so the agent operates a logged-in account.
- 5Ask for a task. Request the action in plain language, naming X or Twitter, the text, and any absolute image paths.
- 6Review before posting. Check the drafted text and the Post button state the agent reports before it publishes.
What the skill makes the agent do
Agent skills are instruction packages that an agent loads when a task matches them. This one covers X (Twitter) work done through a controlled browser rather than the official X API. The skill's source describes verified flows for posting, replying, deleting, and search-and-engage, plus notes on the Draft.js compose editor, which can silently disable the send button if input is not registered.
It also includes stable page selectors, timing guidance (for example waiting after navigation and after the compose box opens), and the daily-reply and job-market-reply playbooks.
- Post a new tweet, with optional local image attachments through the hidden file input
- Reply to posts found in the feed
- Delete, and search-and-engage flows, as described in the skill
- Check the Post button state before clicking
Install the X automation skill
Download the skill ZIP from this page and unzip it. The package contains a single file, SKILL.md.
For a personal install, unzip it into ~/.claude/skills/. For a single project, unzip it into that project's .claude/skills/ folder.
In the Claude apps you can instead upload the ZIP under Settings → Capabilities → Skills.
- Personal: ~/.claude/skills/
- Project: .claude/skills/ inside your project folder
- Claude apps: Settings → Capabilities → Skills, then upload the ZIP
Install the required browser-automation skill first
This skill depends on hive.browser-automation, which supplies the screenshot and coordinate workflow it relies on. The source says to activate browser-automation first and treats the X guidance as an add-on to it.
You also need a browser tool environment that exposes browser_* functions such as navigate, type, click, evaluate and upload, and a logged-in X account inside the controlled browser session.
- hive.browser-automation skill (required)
- Browser tools: navigate, type, click, evaluate, upload
- A logged-in X account in the controlled browser
Trigger the skill with natural requests
Ask for an X task in plain language and mention X or Twitter, so the agent matches the skill. Name the action, the content, and any image file with its absolute path. The skill requires absolute paths, because browser_upload does not expand ~ or accept relative paths.
Examples of requests that should match: 'Post this text on X from my browser session', 'Reply to the posts in my feed with a short answer about our release', 'Attach /home/me/images/chart.png to my next X post'.
- Give the exact text you want posted, or ask the agent to draft it for your review
- Give absolute paths for any local image files
- Say whether you want a single reply or a routine such as the daily-reply playbook
Understand the posting and image flow
For text posts the skill opens the compose box, clicks the text area to put the Draft.js editor into edit mode, types the text, then checks that the Post button is not disabled before clicking it. It advises against clicking blindly.
For images the source is firm: never click the photo button, because it opens the native OS file picker that automation cannot reach. Instead the agent sets the file on the hidden file input with browser_upload, waits for the preview to render, and only then posts.
- Do not click the photo button; set the file on input[data-testid='fileInput']
- Confirm the preview appears before clicking Post
- Allow extra time after posting media
browser_upload(
selector="input[data-testid='fileInput']",
file_paths=["/absolute/path/to/photo.png"],
)Safety, rate limits and what can go wrong
The skill runs scripts and uses the network, and it drives a live logged-in account. Review the SKILL.md and the actions the agent proposes before letting it post. The source notes rate limiting and occasional anti-bot challenges, and warns that automated engagement may violate X's terms and get accounts restricted.
The file is dated 2026-04-11. X's markup changes, so selectors may stop matching. If a step fails, re-check the selectors on the live page before retrying.
- Do not use it for high-volume posting or mass replies
- Expect breakage if X changes its page structure
- For programmatic access, use the official X API instead; this skill does not use it
Tips for reliable runs
Follow the timing guidance in the skill: wait for the page to load and hydrate before querying selectors, and allow about 1.5 seconds for the compose modal to slide in. Text typed in the first moments after opening the composer may be dropped, so verify the text before posting.
Keep test posts in a throwaway account until you have confirmed the flow works for your setup.
- Verify the Post button is enabled before every click
- Close leftover modals with Escape after each action
- Re-verify the flow when X updates its interface
What people use it for
- Drafting and publishing routine posts from a controlled browser session
- Replying to feed posts with a daily-reply routine
- Attaching a local image to a post through the hidden file input
- Building an agent that needs a working reference for X's page structure
Frequently asked questions
What is the X / Twitter Automation agent skill?
It is an agent skill named hive.x-automation that gives an AI agent a tested playbook for operating X through a browser. It covers posting, replying, image attachments and checks on the Post button. It is intended to be used together with the hive.browser-automation skill.
How do I install the hive.x-automation skill?
Download the ZIP from this page and unzip it into ~/.claude/skills/ for personal use or into a project's .claude/skills/ folder. In the Claude apps you can upload the ZIP under Settings → Capabilities → Skills.
What does the skill require?
It requires the hive.browser-automation skill, a browser tool environment exposing browser_* functions, a logged-in X account in the controlled browser session, and absolute paths for local image files.
Does this skill use the official X API?
No. The skill drives the X website with a browser. For API access you should use the official X API instead.
Is the skill free to use?
The source does not state pricing. It is licensed under Apache-2.0, so check that licence text for the terms that apply to using and redistributing it.
Why does my post or reply fail to send?
The Draft.js editor can silently disable the send button when text is not registered. Click the text area again, confirm the Post button is enabled, and verify the text. If the selectors no longer match, X may have changed its markup and the flow needs re-verification.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you are building or running an agent that posts or replies on X from a controlled browser and want a working reference for the page structure and the compose quirks. Pick something else if you need high-volume automation, the official X API, or a tool that will keep working without re-verification as X changes.
Download the skill