← 返回 Skills 市场
man0l

Submit Directories

作者 man0l · GitHub ↗ · v0.1.2 · MIT-0
cross-platform ⚠ suspicious
380
总下载
0
收藏
0
当前安装
3
版本数
在 OpenClaw 中安装
/install submit-directories
功能描述
Use when submitting a product to AI/startup directories - covers the full pipeline from collecting product info, analyzing directories, discovering forms, au...
安全使用建议
This skill appears to do what it says (scan 800+ directories, discover forms, and auto-submit), but review and test before use. Key points: 1) The registry metadata does not declare the many SUBMIT_* environment variables the SKILL.md and code require — be sure to set them intentionally and not in a shared/global environment. 2) Use throwaway email/passwords and confirm the code actually removes or redacts passwords and other PII from any saved JSON or logs (search for any places that write filledDetails or submission_plan.json). 3) The skill will visit lots of external sites and may create accounts or public listings—this can expose product data or assets; inspect directories.json to remove sites you don't trust. 4) It uses Playwright and may prompt you to manually complete OAuth/Google auth or captchas; the gh CLI path will require your GitHub auth and may create forks/PRs. 5) Run the tool in an isolated environment (fresh venv, throwaway .env, non-production credentials) and step through one or two sites manually to confirm behavior before automating wide-scale submissions. If you want higher assurance, ask the author to (a) declare required env vars in registry metadata and (b) add an explicit, audited routine that redacts credentials from all saved files and logs.
功能分析
Type: OpenClaw Skill Name: submit-directories Version: 0.1.2 The skill bundle provides a legitimate automation pipeline for submitting product information to over 800 web directories. The implementation in 'submit_directories.py' and 'discover_forms.py' uses Playwright for browser automation and follows security best practices, such as resolving credentials from environment variables and using CDP-level form filling to prevent data interception by page-level JavaScript. Furthermore, 'submit_directories.py' includes logic to explicitly strip passwords from the 'submission_plan.json' file before saving to disk, and 'SKILL.md' provides clear warnings to users to use throwaway credentials.
能力评估
Purpose & Capability
Name/description match the code: scripts analyze directories.json, discover forms, and auto-fill/submit using Playwright. However the registry metadata lists no required environment variables or primary credential while SKILL.md and the code require many SUBMIT_* environment variables (product fields, email, password, optional GitHub/Twitter). The missing registry declarations are an incoherence to be aware of.
Instruction Scope
SKILL.md and the Python code instruct the agent to visit hundreds of external sites, fill fields, upload assets, and submit entries. The code resolves and fills credentials and other PII from environment vars and logs 'filled' details (filled_log stores values truncated to 50 chars). Although the README warns to use throwaway credentials and claims passwords are stripped before saving, I could not find a guaranteed audit in the provided snippets that sensitive values are always removed from saved JSON/logs—so there is a real risk of accidental credential or PII leakage when the skill writes submission_plan.json or prints filledDetails.
Install Mechanism
No exotic install spec in registry; SKILL.md instructs 'pip install -r requirements.txt' and 'playwright install chromium' and requirements.txt only lists Playwright. No downloads from unknown hosts or archive extraction shown in the package metadata.
Credentials
Requested environment variables (product URL, name, tagline, email, username, password, optional Github/Twitter, asset filenames) are proportionate to the stated purpose. The inconsistency is that the registry metadata declares no required env vars while the runtime clearly depends on many SUBMIT_* variables. The skill also references use of gh CLI for GitHub PR flows (external auth required) which is reasonable but important to know.
Persistence & Privilege
always is false and the skill does not request system-wide persistent privileges. It will run networked browser automation and read/write local files (submission_plan.json, checkpoint.md, directories.json, browser_check_list.json) in the working directory—this is expected for its purpose. Note: autonomous invocation is allowed by default on the platform; combined with broad outgoing network activity this increases blast radius if misused.
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 submit-directories
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /submit-directories 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
v0.1.2
**Password management and GitHub PR process improvements** - Clarified password handling: use environment variable only; passwords are now automatically stripped from `submission_plan.json` before saving. - Updated setup instructions and prompts to reinforce that real passwords should never be used. - Improved documentation for the GitHub PR submission phase: clarified need for `gh` CLI and recommended use of the `SUBMIT_GITHUB_URL` environment variable. - Updated the rules to reflect changes in password management and personal data handling. - Other minor text clarifications.
v0.1.1
**Credentials are now configured via environment variables, with a new `.env.example` provided.** - Added an `.env.example` file to help users set up environment variables for product info and credentials. - Updated documentation to require all credentials/configuration to be set via environment variables, replacing hardcoded values. - Clarified setup instructions for loading local `.env` files safely (with `.gitignore` in place). - Reinforced best practices for secret management and increased safety by discouraging storing sensitive data in scripts.
v0.1.0
Initial release of the submit-directories skill, providing an end-to-end workflow for submitting products to AI and startup directories. - Collects detailed product info interactively and stores submission preferences. - Analyzes over 800 directories for status, authentication, captcha, and submission requirements. - Generates 30 unique copy variations for use in submissions. - Discovers form fields and automates submissions where possible, with clear handling of captchas and OAuth flows. - Supports manual browser-based submission for complex cases and tracks all progress in checkpoint.md. - Comprehensive reporting after each phase; status codes and explanations for all skips and errors.
元数据
Slug submit-directories
版本 0.1.2
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 3
常见问题

Submit Directories 是什么?

Use when submitting a product to AI/startup directories - covers the full pipeline from collecting product info, analyzing directories, discovering forms, au... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 380 次。

如何安装 Submit Directories?

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

Submit Directories 是免费的吗?

是的,Submit Directories 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Submit Directories 支持哪些平台?

Submit Directories 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 Submit Directories?

由 man0l(@man0l)开发并维护,当前版本 v0.1.2。

💬 留言讨论