# Automate Twitter/X Workflows

Teams need repeatable ways to run Twitter/X actions without rebuilding integration steps. This skill provides belt CLI recipes for posting, media publishing, engagement, DMs, follows, deletes, and profile lookups.

## Install

```bash
npx skillstore add inference-sh-skills/twitter-automation
```

## Metadata

- Status: approved
- Slug: inference-sh-skills-twitter-automation
- Skillstore revision: r1
- Version status: missing
- Tree hash: be357a03691fb65e6bc223116009c161e4e1f5e615b7078f9ba3acb342f88801
- Author: inference-sh-skills
- GitHub username: inference-sh-skills
- License: MIT
- Repository: https://github.com/inference-sh-skills/skills/tree/main/tools/social/twitter-automation/
- Ref: a25199bc7d6b82598536822d1738eb5d5f54025b
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: confirmation\_required
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands, network
- Quality score: 38
- Quality tier: warning
- Public page: https://skillstore.pages.dev/skills/inference-sh-skills-twitter-automation
- Manifest: https://skillstore.pages.dev/api/skills/inference-sh-skills-twitter-automation/manifest

## Capabilities

- Provides belt CLI commands for Twitter/X authentication and app execution.
- Shows how to post text updates and create media posts.
- Includes examples for likes, retweets, post deletion, and post lookup.
- Includes examples for sending DMs, following users, and reading user profiles.
- Shows workflows that generate images or videos before publishing them to Twitter/X.

## Use Cases

- Publish Approved Updates: Prepare and run repeatable posting commands for approved text or media updates.
- Share Launch Assets: Generate image or video assets and publish them through a connected Twitter/X account.
- Manage Account Interactions: Retrieve posts or profiles and run approved follow, DM, like, or retweet actions.

## Prompt Templates

### Post a Text Update

```
Help me prepare a safe belt command to post this approved Twitter/X update: [paste text]. Ask before running anything.
```

### Create a Media Post

```
Create a Twitter/X media post workflow using this approved text and media URL. Include a review step before publishing.
```

### Review an Engagement Action

```
Check whether this planned like, retweet, follow, or DM action is appropriate. Then prepare the belt command only if it passes review.
```

### Build a Campaign Workflow

```
Design a human-approved Twitter/X workflow for a launch campaign using posts, media, profile lookups, and engagement actions. Include approval checkpoints and rate limits.
```

## Limitations

- Requires the belt CLI, an inference.sh account, and a configured Twitter/X integration.
- Does not include its own scheduler, application code, or automated retry system.
- Does not add consent checks, moderation checks, or rate limiting by itself.
- Twitter/X API permissions, quotas, and platform rules may limit available actions.

## Best Practices

- Get human approval before publishing, deleting, liking, retweeting, following, or sending DMs.
- Review generated text and media for accuracy, rights, privacy, and platform-policy compliance.
- Keep credentials out of prompts and use the minimum account permissions needed.

## Anti Patterns

- Running bulk follow, like, retweet, or DM loops for audience growth without consent.
- Publishing AI-generated media without reviewing rights, accuracy, or disclosure needs.
- Allowing an agent to delete posts or publish public content without explicit approval.

## Security Audit

- Audited at: 2026-07-05T18:03:08.54\+00:00
- Summary: Most static external-command matches are Markdown code fences or inline app IDs, but several examples do instruct users or agents to run belt and npx commands with real network and account side effects. The highest contextual concern is that the skill enables posting, DMs, follows, likes, retweets, and deletes without visible consent or rate-limit safeguards. No prompt injection attempt was found in the reviewed SKILL.md content.

## Stats

- Views: 56
- Downloads: 5
- Favorites: 0
- Popularity score: 0