← 返回 Skills 市场
seomikewaltman

YoinkIt

作者 seomikewaltman · GitHub ↗ · v1.0.8
cross-platform ⚠ suspicious
852
总下载
3
收藏
3
当前安装
8
版本数
在 OpenClaw 中安装
/install yoinkit
功能描述
Search, analyze, and transcribe content across 13 social platforms — trending topics, video transcripts, post metadata, and multi-platform research workflows.
安全使用建议
Do not install blindly. Key issues: (1) The registry metadata does not list the required YOINKIT_API_TOKEN (and doesn't declare required binaries like curl/jq), but the scripts and SKILL.md clearly require that token and make network requests to https://yoinkit.ai (or a user-set YOINKIT_API_URL). (2) Several example cron jobs instruct the agent to read and update local Obsidian files — the skill does not declare required config paths for that. Before installing, ask the publisher to: publish a homepage/source repo, correct the registry metadata to list YOINKIT_API_TOKEN and required binaries, and explicitly document what local files (if any) the skill will read or write. Verify the yoinkit.ai API domain (and ask for a privacy/data-retention policy). If you proceed, only give the token to the skill after confirming the provider, and avoid scheduling jobs that grant the agent blanket access to sensitive local files or external channels (e.g., Telegram) until you trust the publisher. If you want higher assurance, request the upstream source code repository and confirm that API calls only send the minimal required fields (URLs/ids and not arbitrary local note contents) before enabling automated cron jobs.
功能分析
Type: OpenClaw Skill Name: yoinkit Version: 1.0.8 The OpenClaw AgentSkills skill bundle 'yoinkit' is designed for social media content aggregation via the yoinkit.ai API. All shell scripts (`scripts/*.sh`) correctly sanitize user-provided inputs (URLs, queries, handles) using `jq -sRr @uri` before incorporating them into `curl` commands, effectively preventing command injection. The `SKILL.md` and example cron job configurations (`examples/*.json`) contain instructions for the AI agent that are focused on task execution, output formatting, and specific file interactions (e.g., with an Obsidian vault at a defined path), without any evidence of prompt injection attempts to manipulate the agent for malicious purposes like data exfiltration or unauthorized actions. The skill's behavior is fully aligned with its stated purpose.
能力评估
Purpose & Capability
The skill claims to be an OpenClaw integration for multi-platform social research, which fits the included scripts. However the registry metadata lists no required environment variables or credentials, while the SKILL.md and all scripts clearly require YOINKIT_API_TOKEN (and optionally YOINKIT_API_URL). The metadata also claims no required binaries but the scripts assume standard CLI tools (curl, jq). This mismatch between what the skill actually needs and what the metadata declares is incoherent.
Instruction Scope
The SKILL.md and examples instruct the agent to call the Yoinkit API (network requests) and the example cron jobs include operations to read and update local Obsidian vault files (read/write paths like Research/Daily/YYYY-MM-DD-collection.md). The code files themselves operate on arguments and make API calls, but the examples expect the agent to access local files and save results back — that local file access is not declared in required config paths. The agent could therefore transmit local note contents to the external API or other channels if scheduled, which is beyond what's declared.
Install Mechanism
There is no install spec (instruction-only from a registry perspective), which reduces install-time risk. However the package includes multiple shell scripts that will be executed by the agent runtime. No remote downloads or opaque installers are present. The scripts rely on curl and jq but the registry metadata did not declare these required binaries — a mismatch to be corrected but not inherently high-risk.
Credentials
The skill legitimately needs a single service credential (YOINKIT_API_TOKEN) and optionally YOINKIT_API_URL to point at a server, which is proportionate to its stated purpose. But the registry metadata does not declare any required env vars or primary credential, so the required secret is omitted from the declared surface. Additionally, examples and SKILL.md allow YOINKIT_API_URL to be pointed at localhost or custom endpoints — useful for testing but potentially abused to redirect agent traffic to arbitrary servers if misconfigured.
Persistence & Privilege
The skill is not force-installed (always: false) and uses the normal agent-invocable behavior. The examples include cron jobs (scheduled autonomous runs) and delivery channels (e.g., telegram), which is expected for a research/monitoring skill. This is acceptable, but combined with the above inconsistencies (undeclared env var and instructions to read/write local notes), scheduled autonomous runs increase the impact of any misconfiguration or malicious endpoint.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install yoinkit
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /yoinkit 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.8
- Added note about including the Yoinkit logo (`assets/yoinkit-logo.png`) with results where the platform supports images/media. - Clarified output formatting instructions to send the logo alongside the first result in a conversation. - No changes to commands or API behavior. - Documentation updated only; no changes to code or functionality.
v1.0.7
- Added a new "Output Formatting" section to SKILL.md with best practices for displaying Yoinkit results. - Provides clear guidance on labeling, summarizing, structuring, and formatting output for transcripts, trending, content, and research commands. - Advises on how to handle empty results or unsupported platforms.
v1.0.6
yoinkit 1.0.6 - Documentation section in SKILL.md was removed. - No functional or command changes; updates are documentation-only.
v1.0.5
Aligned all docs, fixed /config syntax, added feed natural language examples
v1.0.4
Added user feed script (8 platforms), Obsidian two-step research workflow, 9 cron examples, aligned all docs
v1.0.3
Improved description, fixed params, shorts URL handling, youtu.be support, YOINKIT_API_URL config
v1.0.2
Fixed params, shorts URL handling, youtu.be support, YOINKIT_API_URL config
v1.0.0
Yoinkit v1.0.0 — Initial Release - Enables transcript extraction and content research from major social platforms. - Provides unified commands for transcript, content, search, trending, and automated research workflows. - Supports multiple platforms including YouTube, TikTok, Instagram, Twitter/X, Reddit, and more. - Offers natural language command capability and detailed configuration instructions. - Requires a Yoinkit API token for access.
元数据
Slug yoinkit
版本 1.0.8
许可证
累计安装 3
当前安装数 3
历史版本数 8
常见问题

YoinkIt 是什么?

Search, analyze, and transcribe content across 13 social platforms — trending topics, video transcripts, post metadata, and multi-platform research workflows. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 852 次。

如何安装 YoinkIt?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install yoinkit」即可一键安装,无需额外配置。

YoinkIt 是免费的吗?

是的,YoinkIt 完全免费(开源免费),可自由下载、安装和使用。

YoinkIt 支持哪些平台?

YoinkIt 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 YoinkIt?

由 seomikewaltman(@seomikewaltman)开发并维护,当前版本 v1.0.8。

💬 留言讨论