← 返回 Skills 市场
Job Watch
作者
ElFilaliHamza
· GitHub ↗
· v1.0.0
· MIT-0
135
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install job-watch
功能描述
Automated job search with 7‑day filtering, scoring, and weekly report. To run the skill send "run job watch now".
使用说明 (SKILL.md)
Job Watch Skill
Automated job search across Moroccan job boards. Filters last 7 days, scores matches against your profile, and produces a structured report.
When to Use
- When you want a weekly automated job search across multiple platforms
- When you need jobs filtered by your specific skills and preferences
- When you want scored job matches delivered as structured reports
External Endpoints
| URL | Purpose | Data Sent |
|---|---|---|
| alwadifa-maroc.com | Job search | Search queries only |
| dreamjob.ma | Job search | Search queries only |
| rekrute.com | Job search | Search queries only |
| emploi-public.ma | Job search | Search queries only |
| anapec.org | Job search | Search queries only |
Security & Privacy
- No credentials required: This skill uses public job board scraping only
- No data persistence: Job data is not stored externally; reports saved locally only
- Rate limiting respected: Respects robots.txt and platform rate limits
- No PII transmission: Your profile data stays in local memory files
Version History
- 1.0.0 (2026-04-03) - Initial release with Moroccan job market focus
Execution Instructions (CRITICAL)
When the user says "run job watch now", DO NOT run any shell commands (including find, ls, cat, bash, python). This skill is designed to be executed entirely by the AI using web-scraping and browser tools, ensuring it works across all OpenClaw configurations without requiring local file system access.
Steps to Execute
- Load user profile from
~/.openclaw/workspace/skills/job-watch/memory/profile.md - Load scoring rules from
~/.openclaw/workspace/skills/job-watch/memory/scoring.md - Load platform list from
~/.openclaw/workspace/skills/job-watch/memory/platforms.md - For each platform, scrape jobs posted in the last 7 days (use browser tools or HTTP requests)
- Filter jobs that match user’s target role, location, and keywords
- Score each job (0–100) based on scoring rules
- Generate report in the format below
- Save report to
~/.openclaw/workspace/skills/job-watch/memory/reports/YYYY-MM-DD-job-watch-report.md
Report Format (copy-paste friendly)
# Job Watch Report – YYYY-MM-DD
## Summary
- Total jobs found: X
- New matches this week: Y
- Average score: Z
## Top Matches (score ≥ 80)
| Score | Title | Company | Platform | Link |
|-------|-------|---------|----------|------|
## Other Matches (score 60–79)
| Score | Title | Company | Platform | Link |
## Low Matches (score \x3C 60) – optional
| Score | Title | Company | Platform | Link |
## Notes / Next Steps
- Apply to top matches by [date]
- Update profile keywords if low matches persist
Example Trigger
User says: "run job watch now"
AI executes the steps above and outputs the report directly in the chat.
安全使用建议
This skill appears to do what it says: read your local profile/scoring/platform files, scrape public Moroccan job boards, score matches, and save a local report. Before using it: (1) Inspect the memory files (profile.md, scoring.md, platforms.md) and remove any sensitive values (passwords, tokens). (2) If you plan to run an installation command from GitHub (README suggests a repo), review that repository’s code — the registry package here is instruction-only, and pulling external code may add behavior not covered in this review. (3) Confirm you are comfortable with scraped results being printed in chat (report content may expose profile-derived data). (4) Ensure scraping targets are public and you comply with their robots.txt and terms; do not provide credentials for sites that require login. If you want higher assurance, request the actual repository/source code (if any) for a code-level review — that would increase confidence above 'medium'.
功能分析
Type: OpenClaw Skill
Name: job-watch
Version: 1.0.0
The 'job-watch' skill is an automated job search tool designed to scrape public Moroccan job boards. It operates by reading local configuration files (profile.md, scoring.md) and explicitly instructs the AI agent to avoid shell commands, relying instead on browser-based scraping. There is no evidence of data exfiltration, malicious execution, or unauthorized access to sensitive system files.
能力评估
Purpose & Capability
Name/description (weekly automated job search, 7‑day filtering, scoring, report) match the SKILL.md steps (scrape listed Moroccan job boards, filter, score, save report). The skill requests no environment credentials or external binaries, and it uses skill-owned memory files (profile.md, scoring.md, platforms.md) which is appropriate for a personalized job search. Minor inconsistency: README suggests an install command that pulls from a GitHub repo, but the registry entry contains no install spec — if you install from that external repo it could introduce code not present in this instruction-only package, so verify the repo before using it.
Instruction Scope
SKILL.md limits execution to web scraping/browser tools (explicitly forbids running shell commands) and instructs the agent to read/write files inside ~/.openclaw/workspace/skills/job-watch/memory/. That scope is coherent for a local job‑watch skill. Two things to note: (1) profile.md contains personal profile details (skills, locations) — the skill states it will not transmit PII, but generating reports in chat will necessarily surface profile-derived matches (local retention vs. what is sent externally should be considered); (2) the prohibition on shell commands is a runtime instruction to the agent but cannot technically prevent other installation methods from adding code that runs on the host — verify what you actually install/run.
Install Mechanism
There is no install spec in the registry (instruction-only), which is low risk because nothing is downloaded or executed by default. However, the README recommends installing via `clawhub install https://github.com/ElFilaliHamza/job-watch-skill` which would fetch external code; that is outside the manifest provided here and should be audited separately before using. No archives, URLs, or extraction steps are present in the registry package itself.
Credentials
The skill requires no environment variables, credentials, or third‑party keys — proportional to its stated purpose of scraping public job boards. The only sensitive data used are the user-provided profile files stored in the skill memory; that is expected but worth protecting (do not include passwords or secrets in profile.md).
Persistence & Privilege
The skill does not request elevated privileges or permanent presence (always:false). Its read/write actions are confined to its own skill workspace memory path (reports written into ~/.openclaw/.../memory/). It does not modify other skills or system-wide config per the provided instructions.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install job-watch - 安装完成后，直接呼叫该 Skill 的名称或使用
/job-watch触发 - 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
v1.0.0
Initial release with Moroccan job board focus:
- Automated weekly job search across major platforms from your choice.
- Filters postings from the past 7 days and matches to user profile
- Scores job matches and produces structured, markdown reports
- No credentials or API required; all reports are saved locally
- Executed fully with browser/HTTP tools—no shell command dependencies
元数据
常见问题
Job Watch 是什么？
Automated job search with 7‑day filtering, scoring, and weekly report. To run the skill send "run job watch now". 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 135 次。
如何安装 Job Watch？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install job-watch」即可一键安装，无需额外配置。
Job Watch 是免费的吗？
是的，Job Watch 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Job Watch 支持哪些平台？
Job Watch 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 Job Watch？
由 ElFilaliHamza（@elfilalihamza）开发并维护，当前版本 v1.0.0。
推荐 Skills