Token导航 LogoToken导航

Agent Skill · 研究检索

id-cv-resume-creator(id 简历 简历创建者)

id-cv-resume-creator 用于查找、检索和筛选相关信息,适合在 OpenClaw 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

id-cv-resume-creator 是研究检索类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为接触敏感数据;许可证为 MIT-0,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 1,431。

总安装 92,763周安装 3,944GitHub Stars 4下载 32,499核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT-0
版本
5.2.4

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:id-cv-resume-creator(id 简历 简历创建者)
来源仓库:https://github.com/rotorstar/id-cv-resume-creator
安装命令:
openclaw skills install id-cv-resume-creator
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install id-cv-resume-creator

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

创建数字身份、简历与履历,支持 PDF、HTML 及交互式 3D 格式输出。适用于求职、项目展示或个人品牌构建的多场景需求。输入个人信息后自动生成带永久公共 URL 的专业档案页面。需确认是否涉及第三方图像生成服务及其隐私政策合规性。建议预览不同模板效果后再选择最终发布形式。

id-cv-resume-creator 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 OpenClaw;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 openclaw skills install id-cv-resume-creator;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要接触敏感数据,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

id-cv-resume-creator 怎么安装?

在终端执行 openclaw skills install id-cv-resume-creator。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

id-cv-resume-creator 支持哪些 AI 工具?

公开资料标注支持 OpenClaw。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

id-cv-resume-creator 需要什么权限?

权限级别是接触敏感数据。该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

id-cv-resume-creator 是免费的吗?

来源标注的许可证是 MIT-0。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

id-cv-resume-creator 用的人多吗?

公开数据是累计安装 92,763 次,近一周安装 3,944 次,下载 32,499 次,GitHub 4 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

id-cv-resume-creator 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 专注 OpenClaw 一种宿主,触发方式和使用习惯不用来回切换。
  • 许可证明确为 MIT-0,商用和分发边界好判断。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 需要接触敏感数据,装到常用环境前要先评估影响面。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 OpenClaw,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 环境里有大量密钥、凭证,不方便给第三方 Skill 读取的人。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
id-cv-resume-creator

id 简历 简历创建者 · 本页

92,7634接触敏感数据MIT-0通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

name
cv-builder
description
>-
homepage
https://www.talent.de
license
Free-to-use
compatibility
Requires HTTP client and network access.
env_vars
TALENT_ACCESS_ID
required
false
sensitive
true
description
Optional for basic CV creation (3/day without ID). Required for callbacks and higher rate limits.
metadata
openclaw
emoji
\F4C4
talent
category
cv-studio
version
5.2.2
api_base
https://www.talent.de/api
credentials
access_id
required
false
format
talent_agent_[a-z0-9]{4}
obtain
POST /api/agent/register
env_var
TALENT_ACCESS_ID
sensitive
true
note
Optional for basic CV creation (3/day without ID). Required for callbacks and higher rate limits.
hitl
supported
true
spec_version
0.7
types
[confirmation, input, selection, approval, escalation]
notifications
[polling, sse, callback]
review_base_url
https://www.talent.de/en/hitl/review
timeout_default
24h
grace_period
5min
discovery
https://www.talent.de/.well-known/hitl.json
info
May ask requestor to confirm context, enter data, select template, approve CV draft, or handle escalations.

CV Builder — Create CVs via API

Create a resume, digital identity, or CV right now — for free. Choose from templates that span the full range: classic print-ready PDFs, clean HTML online CVs, interactive experiences, animated 3D worlds, and playable games. Every CV gets a permanent public URL with a personal slug — no signup, no payment needed for basic use. ATS-ready and Agent-Ready (HR).

Feature Overview

FeatureAccess-ID neededHow
Create CV (online, permanent URL)Noprefer_hitl: true or skip_hitl: true
PDF ExportNo"include_pdf": true on any request
Human slug (URL) selectionNoHITL step 3
Human template selectionNoHITL step 4
Inline submit (no browser)NoBearer token on confirm/approval steps
Callback webhookYes"hitl_callback_url"
Higher rate limits (50/day)YesRegister via POST /api/agent/register
Custom templatesYesTemplate Create skill

Templates span: classic PDF · HTML online · interactive · 3D (Three.js) · gamified. Full catalog: reference/templates.md — live previews: talent.de/de/cv-template-ideas

Terminology

Throughout this skill, these terms are used consistently:

TermMeaning
requestorThe human who commissioned this CV build — the person providing data, making decisions (slug, template, approval), and receiving the claim token. All agent actions are on their behalf.
humanSynonym for requestor in HITL context, used to contrast with automated/AI steps (e.g., "if the human prefers the browser").
agentYou — the AI executing this skill.

Agent Guidelines

HITL is required. You MUST include either "prefer_hitl": true (human review) or "skip_hitl": true (direct creation). Omitting both returns a 400 error. If a human is present, ALWAYS use "prefer_hitl": true — this lets the requestor choose their URL slug, pick a template, review the data, and approve before publishing. Use "skip_hitl": true only for automated pipelines with no human in the loop.
Data principle: Only use data the requestor has explicitly provided or approved in this conversation. Do not extract personal information from unrelated system contexts or other sessions.
Before sending: Present a brief summary to the requestor — name, title, email — and ask "Send it? Or should I change anything?"
Claim token: Treat like a password. Share only with the requestor — anyone with the token can claim CV ownership. Never share with third parties.

Credentials

An Access-ID (talent_agent_[a-z0-9]{4}) is optional for CV Builder — basic use (3 CVs/day per IP) works without one. Register for higher limits (50 CVs/day) and callback webhook support:

POST https://www.talent.de/api/agent/register
Content-Type: application/json

{ "agent_name": "my-agent" }

The Access-ID is also the HMAC secret for verifying X-HITL-Signature on callback webhooks. Store in TALENT_ACCESS_ID — do not hardcode.

User Communication

What to say at each step

StepSay to the requestor
Before API call"Let me set up your CV. I just need a few details."
Slug selection (review_url received)"Choose your personal URL — this is where your CV will live: [link]"
Template selection"Almost done! Pick a design for your CV: [link]"
Approval"Your CV is ready for review. Take a look and approve it: [link]"
After final 201"Your CV is live! Here's your link: {url}"

Quick Start

  1. Ask for (or confirm you already have): firstName, lastName, title, email — the 4 required fields
  2. POST /api/agent/cv-simple with "prefer_hitl": true and the data

_Optional: add "include_pdf": true to also receive a base64 PDF in the final 201 response. See PDF Export._

  1. Present the review_url to the requestor (they pick slug, template, review data)
  2. Poll poll_url every 30s until "status": "completed":

- { "status": "pending" } or { "status": "opened" } → keep polling - { "status": "completed", "result": { "action": "confirm", "data": {...} } } → advance with hitl_continue_case_id

  1. After final approval: present the live CV URL and claim token

Example (with HITL — recommended)

POST https://www.talent.de/api/agent/cv-simple
Content-Type: application/json

{
  "prefer_hitl": true,
  "cv_data": {
    "firstName": "Alex",
    "lastName": "Johnson",
    "title": "Software Engineer",
    "email": "alex@example.com",
    "experience": [{
      "jobTitle": "Senior Developer",
      "company": "Acme Inc.",
      "startDate": "2022-01",
      "isCurrent": true
    }],
    "hardSkills": [{ "name": "React", "level": 4 }],
    "softSkills": [{ "name": "Team Leadership" }],
    "languages": [{ "name": "English", "level": "NATIVE" }]
  }
}

Response (202 — human review required):

{
  "status": "human_input_required",
  "message": "Please confirm: is this CV for you?",
  "hitl": {
    "case_id": "review_a7f3b2c8d9e1f0g4",
    "review_url": "https://www.talent.de/en/hitl/review/review_a7f3b2c8d9e1f0g4?token=abc123...",
    "poll_url": "https://www.talent.de/api/hitl/cases/review_a7f3b2c8d9e1f0g4/status",
    "type": "confirmation",
    "inline_actions": ["confirm", "cancel"],
    "timeout": "24h"
  }
}

Present the review URL to the requestor:

I've prepared your CV. Please review and make your choices here: Review your CV You'll pick your personal URL slug, template design, and approve the final result.

Then poll poll_url until completed, and continue through steps with hitl_continue_case_id. After all steps (confirmation, data review, slug selection, template selection, approval), the final POST returns 201 with the live URL.

Full HITL protocol with all steps, inline submit, edit cycles, and escalation: reference/hitl.md

HITL Multi-Step Flow

The requestor goes through up to 5 review steps. The agent loops: present review URL, poll, continue.

Step 1: Confirmation  →  "For whom is this CV?"
Step 2: Data Review   →  "Are these details correct?"
Step 3: Slug          →  Human picks personal URL slug (e.g. pro, dev, 007)
Step 4: Template      →  Human picks template design
Step 5: Approval      →  Human reviews final CV draft

Each step returns 202. After the requestor decides, continue:

POST https://www.talent.de/api/agent/cv-simple
Content-Type: application/json

{
  "prefer_hitl": true,
  "hitl_continue_case_id": "review_a7f3b2c8d9e1f0g4",
  "slug": "dev",
  "cv_data": { ... }
}
Important: slug and template_id go at the top level of the request, not inside cv_data. When continuing after slug selection, include the human's chosen slug at the top level so the server knows to advance to the template step.

Steps are skipped when you already provide the value:

  • Include slug (top-level) → slug selection step is skipped
  • Include template_id (top-level) → template selection step is skipped
  • Include both → only confirmation, data review, and approval remain

Inline Submit (v0.7)

For simple decisions (confirmation, escalation, approval), the 202 response includes submit_url, submit_token, and inline_actions. Agents can submit directly via Bearer token — ideal for Telegram, Slack, WhatsApp where buttons are supported:

POST {submit_url}
Authorization: Bearer {submit_token}
Content-Type: application/json

{ "action": "confirm", "data": {} }
Always present review_url as a fallback alongside any inline buttons. If the platform does not support buttons (SMS, email, plain text), or the human prefers the browser, they can use the link to complete their decision.

Selection and input types always require the browser (review_url) — they involve complex UI (template grid, data forms). Full inline spec: reference/hitl.md

After the final approval step, submit with hitl_approved_case_id to publish:

POST https://www.talent.de/api/agent/cv-simple
Content-Type: application/json

{
  "hitl_approved_case_id": "review_final_case_id"
}

Response (201):

{
  "success": true,
  "url": "https://www.talent.de/dev/alex-johnson",
  "cv_id": "cv_abc123",
  "claim_token": "claim_xyz789",
  "template_id": "007",
  "quality_score": 65,
  "quality_label": "good",
  "improvement_suggestions": []
}
⚠️ Immediately after receiving a 201 — always share both with the requestor: 1. CV URL: https://www.talent.de/dev/alex-johnson — their live profile 2. Claim link: https://www.talent.de/claim/claim_xyz789 — to take ownership and edit The claim_token is permanent and never expires. Treat it like a password — share only with the requestor.

Present the result:

Your CV is live: talent.de/dev/alex-johnson To claim ownership and edit your CV, visit: talent.de/claim/claim_xyz789 Keep this link safe — it never expires and gives full edit access.

If the response includes improvement_suggestions, share them with the requestor and offer to update the CV:

Your CV score is 35/100. To improve it, I can add: work experience (+25pts), professional summary (+20pts). Want me to ask you a few questions and update it?

Agent Loop (Visual)

Both paths — with and without human review — are shown below. Choose ONE per request.

flowchart TD
    START([Agent starts]) --> CHOICE{prefer_hitl\
or skip_hitl?}

    %% ── skip_hitl path ──────────────────────────────────────────
    CHOICE -->|skip_hitl: true| DIRECT["POST /api/agent/cv-simple\
skip_hitl: true + cv_data"]
    DIRECT --> D201["201 — CV live\
url · claim_token · quality_score\
improvement_suggestions"]
    D201 --> SHARE_NOW["Share url + claim_token\
with requestor immediately!"]
    SHARE_NOW --> QUAL{improvement_suggestions\
present AND attempt < 2?}
    QUAL -->|No / attempt >= 2| DONE_DIRECT([Done])
    QUAL -->|Yes| ASK["Ask requestor the questions\
in each agent_action field"]
    ASK --> REPOST["POST /api/agent/cv-simple\
skip_hitl: true\
enriched cv_data\
(new cv_id each time)"]
    REPOST --> D201

    %% ── prefer_hitl path ─────────────────────────────────────────
    CHOICE -->|prefer_hitl: true| HITL["POST /api/agent/cv-simple\
prefer_hitl: true + cv_data"]
    HITL --> H202["202 human_input_required\
review_url · poll_url · events_url"]
    H202 --> SHOW_URL["Present review_url to requestor\
'Please review here: [link]'"]
    SHOW_URL --> POLL["Poll poll_url every 30s\
(or use events_url for SSE)"]
    POLL --> STATUS{status?}
    STATUS -->|pending / opened\
/ in_progress| POLL
    STATUS -->|completed| ACTION{result.action?}
    STATUS -->|expired| EXP{default_action?}
    STATUS -->|cancelled| CANCELLED([Inform requestor: cancelled])

    EXP -->|skip| AUTO_PUB["CV auto-published\
url from poll status"]
    EXP -->|abort| ABORTED([Inform requestor: expired])

    ACTION -->|confirm / select| CONTINUE["POST cv-simple\
hitl_continue_case_id\
+ ALWAYS include cv_data"]
    ACTION -->|edit| EDIT["Adjust cv_data per note\
then CONTINUE"]
    EDIT --> CONTINUE
    ACTION -->|reject| REJECT["Escalation step\
POST hitl_continue_case_id"]
    REJECT --> H202
    ACTION -->|approve| PUBLISH["POST cv-simple\
hitl_approved_case_id + cv_data"]
    CONTINUE --> H202
    PUBLISH --> DONE_HITL(["201 — CV live\
Share url + claim_token"])

When a HITL case expires, the server applies the default_action configured for that step:

  • skip: CV is auto-published with server-selected slug/template. Poll poll_url — it returns status: "completed" with url in the result.
  • abort: Flow terminates. Inform the requestor the session expired. Start a new HITL flow with prefer_hitl: true if needed.

Quality Improvement Loop (skip_hitl)

After any 201 response the CV is immediately live. Always share url and claim_token with the requestor first.

If improvement_suggestions is non-empty, you may optionally run up to 2 improvement cycles:

  1. For each suggestion: ask the requestor the exact question in agent_action
  2. Re-POST with enriched cv_data + skip_hitl: true
  3. Each re-POST creates a new CV with a new cv_id — the previous one is abandoned
  4. After 2 cycles OR when improvement_suggestions is empty: stop
Do not loop indefinitely. An agent that keeps re-posting creates duplicate CVs and wastes the requestor's time.

For prefer_hitl flows: after approval, the 201 is already quality-assessed (human reviewed the data). If suggestions appear, ask the requestor directly — do not restart the HITL flow.

Direct Creation (No Human Present)

For fully automated pipelines, batch operations, or when the requestor explicitly says "just create it" — set "skip_hitl": true:

POST https://www.talent.de/api/agent/cv-simple
Content-Type: application/json

{
  "skip_hitl": true,
  "cv_data": {
    "firstName": "Alex",
    "lastName": "Johnson",
    "title": "Software Engineer",
    "email": "alex@example.com"
  }
}

Response (201):

{
  "success": true,
  "url": "https://www.talent.de/pro/alex-johnson",
  "cv_id": "cv_abc123",
  "claim_token": "claim_xyz789",
  "template_id": "018",
  "hitl_skipped": true,
  "quality_score": 20,
  "quality_label": "basic",
  "improvement_suggestions": [
    {
      "field": "experience",
      "issue": "No work experience — CV has low ATS compatibility",
      "agent_action": "Ask: 'What positions have you held? Part-time and internships count.'",
      "impact": "+25 quality points",
      "priority": "high"
    }
  ],
  "next_steps": "Share improvement_suggestions with the requestor and ask the questions in agent_action. Then update the CV via POST /api/agent/cv-simple...",
  "auto_fixes": []
}

In direct mode, the server auto-assigns slug (pro by default) and template (018 Amber Horizon). The requestor has no choice. Use this only when no human needs to review.

You MUST choose one: "prefer_hitl": true or "skip_hitl": true. Omitting both returns a 400 error.

All fields beyond the 4 required ones are optional. Omit what you don't have — don't send empty arrays.

PDF Export

Get a downloadable PDF alongside the CV. Three visual themes available:

ThemeStyleBest for
classicSingle-column, red accent (default)Traditional industries
modernTwo-column sidebar, blue accentTech & creative roles
minimalMonochrome, generous whitespaceExecutive & senior roles

Option A: Inline with CV creation

Add "include_pdf": true to your request. The response includes a base64-encoded PDF:

POST https://www.talent.de/api/agent/cv-simple
Content-Type: application/json

{
  "prefer_hitl": true,
  "include_pdf": true,
  "pdf_format": "A4",
  "pdf_theme": "modern",
  "cv_data": {
    "firstName": "Alex",
    "lastName": "Johnson",
    "title": "Software Engineer",
    "email": "alex@example.com"
  }
}

Response includes a pdf object:

{
  "success": true,
  "url": "https://www.talent.de/pro/alex-johnson",
  "cv_id": "cv_abc123",
  "claim_token": "claim_xyz789",
  "pdf": {
    "base64": "JVBERi0xLjQK...",
    "size_bytes": 6559,
    "generation_ms": 226,
    "format": "A4"
  }
}

Option B: Generate PDF for existing CV

POST https://www.talent.de/api/agent/cv/pdf
Content-Type: application/json

{
  "cv_id": "cv_abc123",
  "format": "A4",
  "theme": "minimal"
}

Returns the PDF binary directly (Content-Type: application/pdf). Format options: A4 (default), LETTER. Theme options: classic (default), modern, minimal.

PDF generation takes ~200ms (no headless browser needed).

Server-Side Intelligence

You do not need to check slug availability or validate data — the server handles it:

  • Slug uniqueness: A slug is NOT globally unique — it's unique per person. pro/thomas-mueller and pro/anna-schmidt can coexist. Only the combination of slug + firstName + lastName is reserved. That's why the server needs the name first to check availability.
  • Slug auto-selection: If omitted (and no HITL), the server picks pro. If that slug is already in use for this person's name, it tries the next available slug automatically. In HITL mode, the human can choose their slug interactively. Popular picks (excerpt): 007 · 911 · dev · api · pro · gpt · web · ceo · cto · ops · f40 · gtr · amg · gt3 · zen · art · lol · neo · 404 · 777. Full list: GET /api/public/slugs
  • Template default: 018 (Amber Horizon) if omitted.
  • Date normalization: 2024 becomes 2024-01-01, 2024-03 becomes 2024-03-01.
  • Language levels: Normalized to CEFR (NATIVE, C2, C1, B2, B1, A2, A1).
  • Human-readable errors: If something goes wrong, the response explains what to fix in plain English.
  • Auto-fix summary: The auto_fixes array tells you what the server adjusted (e.g. "Slug 'pro' is already in use for this name, using 'dev' instead").

Skills Use 4 Arrays

  • hardSkills — technical skills, optional level 1-5
  • softSkills — name only
  • toolSkills — name only
  • languages — with CEFR level: NATIVE, C2, C1, B2, B1, A2, A1

Do not use a generic skills array — it will be ignored.

Common Mistakes

WrongCorrectWhy
"role": "Engineer""jobTitle": "Engineer"Experience uses jobTitle, not role or position
"start": "2022" / "end": "2023""startDate": "2022-01" / "endDate": "2023-06"Wrong field names — start/end are silently ignored in experience and education
"skills": [...]"hardSkills": [...] etc.Generic skills array is ignored — use 4 separate arrays
"slug": "dev" inside cv_data"slug": "dev" at top levelslug and template_id are request-level fields, not inside cv_data
"startDate": "January 2024""startDate": "2024-01"Dates must be YYYY or YYYY-MM format
Sending empty arrays "hobbies": []Omit the field entirelyDon't send empty arrays — omit what you don't have
POST /respond without Authorization headerWait for status=completed via poll_urlInline submit requires submit_token from the 202 response hitl object
POST /respond on an approval type stepPoll until completed, then hitl_approved_case_idApproval steps require browser review — inline submit not permitted
hitl_continue_case_id without cv_dataAlways include the full cv_data objectThe server needs cv_data on every POST to build the next review step UI
skip_hitl: true when a HITL chain is ongoingContinue the chain with hitl_continue_case_idCreates a separate CV and bypasses the human's review — share one chain per CV

Guardrails

  • Rate limits (with Access-ID): 50 CVs/day
  • Rate limits (without Access-ID): 3 CVs/day per IP
  • Never auto-commit without requestor approval
  • Claim tokens are permanent — treat as passwords

References

Specs

平台分布

OpenClaw

78.71%
按下载量换算25,580

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

未展示

权限和风险

接触敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills