Hubbestaiskill.com

Home / All Skills / Content & Media Creation / twitter-automation

Content & Media Creation · 101-skills/superpowers

twitter-automation

Automate Twitter/X with posting, engagement, and user management via inference.sh CLI. Apps: x/post-tweet, x/post-create (with media), x/post-like, x/post-retweet, x/dm-send, x/user-follow. Capabilities: post tweets, schedule content, like posts, retweet, send DMs, follow users, get profiles. Use for: social media automation, content scheduling, engagement bots, audience growth, X API. Triggers: twitter api, x api, tweet automation, post to twitter, twitter bot, social media automation, x automation, tweet scheduler, twitter integration, post tweet, twitter post, x post, send tweet

L4 High-Privilege · Graded auth & full auditV1 Commit-PinnedExecute CommandNetwork AccessWrite FileDual-pass Audit

Author 101-skillsGitHub 101-skills/superpowers ↗License MIT（仓库 README 徽标 'License: MIT' 与 .claude-plugin/plugin.json 中 "license": "MIT" 声明；但 pin 树内无 LICENSE 文件，GitHub API 对两 fork 亦报无 license —— 授权只出现在元数据/徽标，无正式许可文本）Commit becc256497

Control levels describe skill capabilities and suggested control measures; they do not represent quality or safety certifications, nor do they dictate usability. Usage decisions rest entirely with users and administrators. Different agent hosts apply varying constraints; do not assume hosts provide automatic protection. Third-party skill copyrights belong to their original authors. Use at your own risk.

1How It Works & Mechanism

The implementation mechanism relies on a 'whitelisted CLI template set' rather than local code: the frontmatter restricts host shell execution permissions solely to belt (the inference.sh CLI). The SKILL.md body consists entirely of command templates in the format of 'belt app run x/… --input …', where the agent populates the user's requirements into the --input JSON, and the host Bash executes it on its behalf.

tools/social/twitter-automation/SKILL.md

allowed-tools: Bash(belt *)

The allowed-tools in the frontmatter represents a generic constraint for the tools/* series in this repository: the model cannot execute arbitrary shell commands and is confined strictly to the belt command family, reducing the execution surface to a single external CLI.

Capability boundary = 'Available Apps' catalog: 8 platform apps under x/* (post tweet, media post, like, retweet, delete post, get post, send DM, follow user, get user) constitute all performable actions. The actual X API implementation (OAuth, network requests, rate limiting) is handled on the inference.sh platform side; this skill only sees app-ids and JSON parameters.

tools/social/twitter-automation/SKILL.md

| Post Tweet | `x/post-tweet` | Post text tweets |

The app-ids in the catalog lines serve as the executable units. The skill ships with no X API code of its own and cannot function without belt or an authenticated session—it acts essentially as an 'index card for platform capabilities'.

Invocation is automatically triggered by the model: the description embeds a broad vocabulary of Triggers (twitter api / tweet automation / post to twitter / x automation / tweet scheduler …). This skill does not declare any 'manual-only' flags such as disable-model-invocation, meaning it can be automatically invoked whenever related phrases appear in the conversation.

tools/social/twitter-automation/SKILL.md

Triggers: twitter api, x api, tweet automation, post to twitter

The trigger words even include 'tweet scheduler', whereas the main text provides no scheduling capabilities (see second_pass discrepancies)—making the trigger surface broader than the actual capability scope.

'Generate → Publish' workflows are implemented through two-step chaining: first, run an image/video generation app to obtain an output URL (demonstrated using shell redirection > image.json to write to disk), then pass the URL as media_url to x/post-create to publish a media post.

tools/social/twitter-automation/SKILL.md

belt app run falai/flux-dev-lora --input '{"prompt": "sunset over mountains"}' > image.json

This workflow highlights that the skill is designed not as a standalone tool, but as a stage in a content pipeline; the > redirection implies that the agent will write temporary JSON files to the local working directory when executing the example.

Interconnected ecosystem bootstrapping: the Related Skills section uses `npx skills add inference-sh/skills@<skill>` to guide the installation of underlying or sibling skills (full infsh-cli, ai-image-generation, ai-video-generation, ai-avatar-video), directly embedding package installation instructions into the prompt provided to the model.

tools/social/twitter-automation/SKILL.md

npx skills add inference-sh/skills@ai-video-generation

npx skills add is the installation command for the skills.sh ecosystem. Skills reference each other via package managers, forming a skill family centered around the inference.sh platform.

2Core Capabilities

01Publish text-only posts (x/post-tweet)

02Publish posts with media (x/post-create, passing image/video URLs via media_url)

03Engagement: liking and retweeting (x/post-like, x/post-retweet)

04Account management including DMs and following (x/dm-send, x/user-follow, x/user-get)

05Delete and retrieve posts by ID (x/post-delete, x/post-get)

06'Generate → Publish' pipelines chained with image/video generation skills (two workflow examples)

3External Dependencies

Kind	Dependency / Endpoint
cli	belt（inference.sh CLI）
package	belt-sh/cli（belt CLI 的 skill 包，npx skills add 安装）
package	inference-sh/skills 相关技能包（npx skills add inference-sh/skills@…）
network	inference.sh 平台（x/* app 执行与文档）
network	安装说明托管（cli-install.md）
network	横幅示例图（cloud.inference.sh 静态文件）

4Recommended Governance & Evidence

Suggested Controls for Safe Operation

For Agent Runtime

Graded/per-use authorization, sandbox recommended

For Administrators

Approval workflow, rollback support, full audit

Evidence & Triggers (File Paths & Excerpts)

网络端点: Declares external network access: platform belt CLI → inference.sh platform (x/* apps invoke the X API platform-side) ['inference.sh', '
本地工作目录: File I/O and artifact output: local side effects in example commands: `> image.json` / `> video.json` shell redirection, and `belt app sample x/p
SKILL.md: Declares invocation of host CLI interpreter or external runtime environment

Audit Notes: Contains no bundled scripts, credential harvesting, or direct network implementation code; all actions are consolidated into host execution of a single official CLI `belt` (whitelisted in frontmatter), with predictable egress targets (the inference.sh platform, which then calls the X API). Key considerations: ① Performs write operations on the user's real X account (posting/DM/following/deleting), some of which are irreversible and publicly visible; ② Consumes inference.sh account credits, which are billable on the platform side; ③ The claimed schedule capability in the description has no corresponding implementation in the source code; ④ Supply chain: belt CLI and related skill packages are distributed via npx/npm from third-party repositories without version pinning, and this repo is an unrenamed mirror of inference-sh/skills without a tree-level LICENSE file. According to unified classification criteria: `belt` invoked by this skill is distributed via third-party npm/npx (installed via `npx skills add belt-sh/cli`), runs under authenticated `belt login` state, and consumes inference.sh account quota/credits—falling under 'invoking third-party CLI with quota/auth touchpoints', hence graded Orange (Use with Evaluation); the skill source code itself contains no token/secret extraction.

5Second-Pass Verification

[discrepancy] description 声称支持 schedule content / content scheduling / tweet scheduler — grep 全文：schedul* 仅出现在第 3 行 description 字符串内；Available Apps 表 9 个 app 与全部示例均无任何定时/调度命令。真正调度需宿主外部定时机制，本 skill 无法兑现该宣称。
[ok] post/like/retweet/DM/follow/get/delete 能力与 app 表一一对应 — Apps 表含 x/post-tweet、x/post-create、x/post-like、x/post-retweet、x/post-delete、x/post-get、x/dm-send、x/user-follow、x/user-get，各能力均有对应示例。
[ok] 执行面仅 belt CLI（allowed-tools: Bash(belt *)） — frontmatter 白名单存在且为唯一授权；全文无其他可执行代码/脚本文件。
[ok] 无直连网络调用与凭证文本 — curl/wget/fetch/api_key/bearer/token/secret 扫描零命中（token 一词仅出现在外部文档语境之外——本文件无）。登录态由 belt login 管理（'belt login' 第 19 行）。
[ok] 网络端点仅文档/横幅/安装说明 — 出现的 URL 只有 inference.sh/docs（文档）、raw.githubusercontent.com（cli-install.md）、cloud.inference.sh（横幅图）；x/* app 的真实 API 端点在 belt 与平台侧，源码内不可见。
[ok] 本地文件写入面 — 仅示例命令的 shell 重定向（> image.json、> video.json）与 belt app sample --save input.json，属用户/agent 执行示例时的行为，非 skill 代码主动写文件。
[ok] license/stars/commit 元数据 — 两 fork 本地 HEAD == pin becc25649700（2026-08-03）；GitHub API：101-skills/superpowers stars=0、pushed_at 2026-08-31T22:58:53Z、license=None；树内无 LICENSE 文件；plugin.json 声明 MIT 且 repository 指向 inference-sh/skills。
[unlocatable] installs 安装量 — GitHub API 无安装量字段，skills.sh 生态未提供公开计数，置 null。

6Verdict & Scenarios

Condenses 'X automation' into a compact catalog card: 9 apps with pre-built command templates, enabling the agent to execute common account operations without understanding the X API.

Deliberately narrowed permission scope: the allowed-tools: Bash(belt *) whitelist requires the host to permit only a single CLI, minimizing the attack surface of arbitrary shell execution by the model.

Comes with built-in two-step 'Generate → Publish' workflow examples, directly integrating with image/video generation skills in the same repository to form a content automation pipeline.

Comprehensive documentation links (X.com Integration setup page), reducing onboarding friction for users.

Best for: Ideal for users already registered with inference.sh who accept third-party platform management of their X integration in content operations or marketing automation scenarios: chaining 'AI-generated assets → post to X' into an agent workflow; also suitable as the publishing execution layer for larger automation systems (e.g., scheduled content systems).

Not for: Unsuitable for users reluctant to authorize their X accounts to third-party platforms or those concerned with granular X API compliance details (custom OAuth apps, rate limits, audits); unsuitable for scenarios requiring true scheduling capabilities (this skill does not implement scheduling); not for offline or zero-cost execution; not for fully unattended bulk posting without human review.

Install Agent-Ready

① Verified Mirror 2026-09-09

Option A · Direct DownloadDownload twitter-automation.tar.gz

sha256: 21d96a6cedba3a4e…

Option B · Copy Guide for Agent

Agents automatically download and verify mirrors from this JSON guide.

② Upstream GitHub

GitHub Repo ↗

Pinned to commit becc256497.