← 返回 Skills 市场
capsolver

CapSolver

作者 capsolver · GitHub ↗ · v1.0.1 · MIT-0
cross-platform ⚠ suspicious
309
总下载
1
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install capsolver-captcha-solver
功能描述
Use capsolver to automatically resolve Geetest, reCAPTCHA v2, reCAPTCHA v3, MTCaptcha, DataDome, AWS WAF, Cloudflare Turnstile, and Cloudflare Challenge, etc.
使用说明 (SKILL.md)

CapSolver Skill

Use this skill to automatically resolve various CAPTCHA challenges using the CapSolver API service. The skill supports Geetest, reCAPTCHA v2/v3, MTCaptcha, DataDome, AWS WAF, Cloudflare Turnstile, Cloudflare Challenge, and more.

Authentication: Set API_KEY in your environment or in a .env file in the repo root.

Errors: If it fails, the script will exit with code 1.

Solver

Task(Recognition)

ImageToTextTask

Solve text-based captcha.

Command:

python3 ./scripts/solver.py ImageToTextTask --body "base64_image_data"
python3 ./scripts/solver.py ImageToTextTask --body "base64_image_data" --module "module_001"

Optional:

ReCaptchaV2Classification

Classify reCAPTCHA v2 images.

Command:

python3 ./scripts/solver.py ReCaptchaV2Classification --question "question" --image "base64_image_data"

Optional:

AwsWafClassification

Classify AWS WAF images.

Command:

python3 ./scripts/solver.py AwsWafClassification --question "question" --images "base64_image_data1" "base64_image_data2" "base64_image_data3"

Too many images may exceed the command line length limit. You can try writing the base64 values of the images line by line to a file (such as aws_images.txt) and then using the xargs command to pass them to the --images parameter:

cat aws_images.txt | xargs python3 ./scripts/solver.py AwsWafClassification --question "question" --images

Optional:

VisionEngine

Advanced AI vision-based captcha solving.

Command:

python3 ./scripts/solver.py VisionEngine --module "module" --image "base64_image_data" --imageBackground "base64_image_background_data"

Optional:

Task(Token)

GeeTest

Solve GeeTest captcha (v3/v4).

Command:

python3 ./scripts/solver.py GeeTestTaskProxyLess --websiteURL "https://example.com/" --captchaId "captcha_id"

Optional:

  • --websiteURL: Web address of the website using geetest (Ex: https://geetest.com).
  • --gt: Only Geetest V3 is required.
  • --challenge: Only Geetest V3 is required.
  • --captchaId: Only Geetest V4 is required.
  • --geetestApiServerSubdomain: Special api subdomain, example: api.geetest.com.
  • --max-retries: Maximum number of retries (default: 60).

reCAPTCHA v2

Solve Google reCAPTCHA v2 (checkbox/invisible).

Command:

python3 ./scripts/solver.py ReCaptchaV2TaskProxyLess --websiteURL "https://example.com" --websiteKey "site_key"
python3 ./scripts/solver.py ReCaptchaV2Task --websiteURL "https://example.com" --websiteKey "site_key" --proxy "host:port:username:password"

Optional:

  • --websiteURL: The URL of the target webpage that loads the captcha, It’s best to submit the full URL instead of just the host.
  • --websiteKey: Recaptcha website key.
  • --proxy: Learn Using proxies: https://docs.capsolver.com/guide/api-how-to-use-proxy/.
  • --pageAction: For ReCaptcha v2, if there is an sa parameter in the payload of the /anchor endpoint, please submit its value.
  • --enterprisePayload: For ReCaptchaV2 enterprise version, if there is an s parameter in the payload of the /anchor endpoint, please submit its value.
  • --isInvisible: Pass true if there is no “I’m not a robot” checkbox but the challenge will still appear, usually required in v2 invisible mode.
  • --isSession: Session mode, when enabled, will return a recaptcha-ca-t value, which is used as a cookie. It usually appears in v3. Note: Some websites require a recaptcha-ca-e value, which usually appears in v2. If this value is present, it will be automatically returned without any additional parameter settings.
  • --max-retries: Maximum number of retries (default: 60).

reCAPTCHA v3

Solve Google reCAPTCHA v3.

Command:

python3 ./scripts/solver.py ReCaptchaV3TaskProxyLess --websiteURL "https://example.com" --websiteKey "site_key"
python3 ./scripts/solver.py ReCaptchaV3Task --websiteURL "https://example.com" --websiteKey "site_key" --proxy "host:port:username:password"

Optional:

  • --websiteURL: The URL of the target webpage that loads the captcha, It’s best to submit the full URL instead of just the host.
  • --websiteKey: Recaptcha website key.
  • --proxy: Learn Using proxies: https://docs.capsolver.com/guide/api-how-to-use-proxy/.
  • --pageAction: For ReCaptcha v3, You can find the value of the action parameter by searching for grecaptcha.execute.
  • --enterprisePayload: For the enterprise version, search for grecaptcha.enterprise.render and pass the s parameter.
  • --isSession: Session mode, when enabled, will return a recaptcha-ca-t value, which is used as a cookie. It usually appears in v3. Note: Some websites require a recaptcha-ca-e value, which usually appears in v2. If this value is present, it will be automatically returned without any additional parameter settings.
  • --max-retries: Maximum number of retries (default: 60).

MTCaptcha

Solve MTCaptcha.

Command:

python3 ./scripts/solver.py MtCaptchaTaskProxyLess --websiteURL "https://example.com" --websiteKey "site_key"
python3 ./scripts/solver.py MtCaptchaTask --websiteURL "https://example.com" --websiteKey "site_key" --proxy "host:port:username:password"

Optional:

DataDome

Solve DataDome.

Command:

python3 ./scripts/solver.py DatadomeSliderTask --captchaUrl "https://geo.captcha-delivery.com/xxxxxxxxx" --userAgent "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/146.0.0.0 Safari/537.36" --proxy "host:port:username:password"

Optional:

  • --captchaUrl: If the url contains t=bv that means that your ip must be banned, t should be t=fe.
  • --userAgent: It needs to be the same as the userAgent you use to request the website. Currently we only support the following userAgent.
  • --proxy: Learn Using proxies: https://docs.capsolver.com/guide/api-how-to-use-proxy/.
  • --max-retries: Maximum number of retries (default: 60).

AWS WAF

Solve AWS WAF.

Command:

python3 ./scripts/solver.py AntiAwsWafTask --websiteURL "https://example.com" --awsChallengeJS "https://path/to/challenge.js" --proxy "host:port:username:password"
python3 ./scripts/solver.py AntiAwsWafTaskProxyLess --websiteURL "https://example.com" --awsChallengeJS "https://path/to/challenge.js"
python3 ./scripts/solver.py AntiAwsWafTaskProxyLess --websiteURL "https://example.com"

Optional:

  • --websiteURL: If the url contains t=bv that means that your ip must be banned, t should be t=fe.
  • --proxy: Learn Using proxies: https://docs.capsolver.com/guide/api-how-to-use-proxy/.
  • --awsKey: The key value returned by the captcha page.
  • --awsIv: The iv value returned by the captcha page.
  • --awsContext: The context value returned by the captcha page.
  • --awsChallengeJS: The challenge.js link returned by the captcha page.
  • --awsApiJs: The jsapi.js link returned by the captcha page.
  • --awsProblemUrl: The problem endpoint url containing keywords like problem, num_solutions_required, etc..
  • --awsApiKey: The api_key value of the problem endpoint.
  • --awsExistingToken: The aws-waf-token used for the last verification.
  • --max-retries: Maximum number of retries (default: 60).

Cloudflare Turnstile

Solve Cloudflare Turnstile.

Command:

python3 ./scripts/solver.py AntiTurnstileTaskProxyLess --websiteURL "https://example.com" --websiteKey "site_key"

Optional:

  • --websiteURL: The address of the target page.
  • --websiteKey: Turnstile website key.
  • --action: The value of the data-action attribute of the Turnstile element if it exists.
  • --cdata: The value of the data-cdata attribute of the Turnstile element if it exists.
  • --max-retries: Maximum number of retries (default: 60).

Cloudflare Challenge

Solve Cloudflare Challenge (5-second shield).

Command:

python3 ./scripts/solver.py AntiCloudflareTask --websiteURL "https://example.com" --proxy "host:port:username:password"

Optional:

  • --websiteURL: The address of the target page.
  • --proxy: Learn Using proxies: https://docs.capsolver.com/guide/api-how-to-use-proxy/.
  • --userAgent: The user-agent you used to request the target website. Only Chrome’s userAgent is supported.
  • --html: The response of requesting the target website, it usually contains "Just a moment…" and status code is 403. we need this html for some websites, please be sure to use your sticky proxy to dynamically scrape the HTML every time.
  • --max-retries: Maximum number of retries (default: 60).

Response example

Output: All commands return JSON objects with task-specific solution fields.

ImageToTextTask

{
  "errorId": 0,
  "errorCode": "",
  "errorDescription": "",
  "status": "ready",
  "solution": {
    "text": "44795sds",
    // number module:
	"answers": ["100", "1330", "147", "248", "303", "439", "752", "752", "752"],
  },
  "taskId": "..."
}

ReCaptchaV2Classification

multi objects:

{
 "errorId": 0,
 "status": "ready",
 "solution": {
   "type": "multi",
   "objects": [
     0,
     1,
     2,
     3
   ],
   "size": 4,
   // 3 or 4
 },
 "taskId": "cbb1c730-e569-4ba6-b5fc-e06377694aa7"
}

single object:

{
 "errorId": 0,
 "status": "ready",
 "solution": {
   "type": "single",
   "hasObject": true,
   "size": 1,
 },
 "taskId": "cbb1c730-e569-4ba6-b5fc-e06377694aa7"
}

AwsWafClassification

{
 "errorId": 0,
 "status": "ready",
 "solution": {
   //carcity point
   "box": [
     116.7,
     164.1
   ],
   // grid type, objects means the image index that matches the question
   "objects": [0, 1, 3, 4, 6],
   //if question include `bifurcatedzoo`
   "distance": 500
 },
 "taskId": "cbb1c730-e569-4ba6-b5fc-e06377694aa7"
}

VisionEngine

{
  "errorId": 0,
  "errorCode": "",
  "errorDescription": "",
  "status": "ready",
  "solution": {
     "distance": 213,
  },
  "taskId": "cbb1c730-e569-4ba6-b5fc-e06377694aa7"
}

GeeTest

Geetest v3:

{
  "errorId": 0,
  "taskId": "e0ecaaa8-06f6-41fd-a02e-a0c79b957b15",
  "status": "ready",
  "solution": {
    "challenge": "...",
    "validate": "...",
    "seccode": "...",
    "userAgent": "..."
  },
}

Geetest v4:

{
  "errorId": 0,
  "taskId": "e0ecaaa8-06f6-41fd-a02e-a0c79b957b15",
  "status": "ready",
  "solution": {
    "captcha_id": "",
    "captcha_output": "",
    "gen_time": "",
    "lot_number": "",
    "pass_token": "",
    "risk_type": "slide"
  }
}

reCAPTCHA

reCAPTCHA v2/v3:

{
    "errorId": 0,
    "errorCode": null,
    "errorDescription": null,
    "solution": {
        "userAgent": "xxx", // User-Agent
        "secChUa": "xxx", // Sec-Ch-Ua
        "createTime": 1671615324290, // The creation time of the token
        "gRecaptchaResponse": "3AHJ......", // token
        "recaptcha-ca-t": "AbEM......", // Some v3 websites have session mode. After enabling isSession, this parameter will be returned and used as a cookie.
        "recaptcha-ca-e": "Abp_......" // Some v2 websites have this parameter, which is used as a cookie. If there is such a value, it will be automatically returned.
    },
    "status": "ready"
}

MTCaptcha

{
  "errorId": 0,
  "taskId": "646825ef-9547-4a29-9a05-50a6265f9d8a",
  "status": "ready",
  "solution": {
    "token": ""
  }
}

DataDome

{
  "errorId": 0,
  "errorCode": null,
  "errorDescription": null,
  "solution": {
    "cookie": "datadome=yzj_BK...S0; Max-Age=31536000; Domain=; Path=/; Secure; SameSite=Lax"
  },
  "status": "ready"
}

AWS WAF

{
  "errorId": 0,
  "taskId": "646825ef-9547-4a29-9a05-50a6265f9d8a",
  "status": "ready",
  "solution": {
    "cookie": "223d1f60-0e9f-4238-ac0a-e766b15a778e:EQoAf0APpGIKAAAA:AJam3OWpff1VgKIJxH4lGMMHxPVQ0q0R3CNtgcMbR4VvnIBSpgt1Otbax4kuqrgkEp0nFKanO5oPtwt9+Butf7lt0JNe4rZQwZ5IrEnkXvyeZQPaCFshHOISAFLTX7AWHldEXFlZEg7DjIc="
  }
}

Cloudflare Turnstile

{
  "errorId": 0,
  "taskId": "61138bb6-19fb-11ec-a9c8-0242ac110006",
  "status": "ready",
  "errorCode": null,
  "errorDescription": null,
  "solution": {
    "token": "0.mF74FV8wEufAWOdvOak_xFaVy3lqIDel7SwNhw3GgpICSWwTjYfrQB8mRT1dAJJBEoP7N1sESdp6WH9cTS1T0catWLecG3ayNcjwxVtr3hWfS-dmcBGRTx4xYwI64sAVboYGpIyuDBeMIRC3W8dK35v1nDism9xa595Da5VlXKM7hk7pIXg69lodfiftasIkyD_KUGkxBwxvrmz7dBo10-Y5zvro9hD4QKRjOx7DYj9sumnkyYCDx0m4ImDIIkNswfVTWI2V22wlnpHdvMgdtKYgOIIAU28y9gtdrdDkpkH0GHcDyd15sxQGd9VjwhGZA_mpusUKMsEoGgst2rJ3zA.UWfZupqLlGvlATkPo3wdaw.38d55cd0163610d8ce8c42fcff7b62d8981495cc1afacbb2f14e5a23682a4e13",
    "type": "turnstile",
    "userAgent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/108.0.0.0 Safari/537.36"
  }
}

Cloudflare Challenge

{
  "errorId": 0,
  "taskId": "df944101-64ac-468d-bc9f-41baecc3b8ca",
  "status": "ready",
  "errorCode": "",
  "errorDescription": "",
  "solution": {
    "cookies": {
        "cf_clearance": "Bcg6jNLzTVaa3IsFhtDI.e4_LX8p7q7zFYHF7wiHPo...uya1bbdfwBEi3tNNQpc"
    },
    "token": "Bcg6jNLzTVaa3IsFhtDI.e4_LX8p7q7zFYHF7wiHPo...uya1bbdfwBEi3tNNQpc",
    "userAgent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/139.0.0.0 Safari/537.36"
  }
}
安全使用建议
This skill appears to implement a legitimate CapSolver client (it posts tasks to api.capsolver.com and needs API_KEY). Before installing: 1) Verify the skill source/provenance (homepage is capsovler.com but the registry owner/slug are unfamiliar); prefer official or well-known vendor-published packages. 2) Correct the manifest mismatch: the registry metadata should declare API_KEY as required — don't trust the registry 'none' claim. 3) Only provide the API_KEY you intend to use for CapSolver and avoid passing unrelated credentials as CLI args. 4) Run the script in an isolated environment (container or VM) and inspect the code yourself if you will provide real secrets. 5) Consider legal/ethical implications of bypassing CAPTCHAs for your use case. If you want higher assurance, ask the publisher to fix the metadata inconsistency and to publish the package from an authenticated, known repository (e.g., official GitHub/org).
功能分析
Type: OpenClaw Skill Name: capsolver-captcha-solver Version: 1.0.1 The skill bundle is a legitimate integration for the CapSolver CAPTCHA solving service. The primary script, scripts/solver.py, functions as a CLI wrapper that sends CAPTCHA data to the official CapSolver API (api.capsolver.com) and retrieves solutions. The code follows standard practices, using argparse for input and requests for network communication, with no evidence of malicious execution, data exfiltration, or prompt injection in SKILL.md.
能力评估
ℹ Purpose & Capability
The skill's name/description (captcha solving) matches the included code and runtime instructions: the script uses requests to call https://api.capsolver.com and implements many captcha task types. Requiring an API key and optional proxy parameters is expected for this functionality. However, the registry metadata earlier claimed 'Required env vars: none' while SKILL.md and scripts clearly require API_KEY — a manifest inconsistency.
✓ Instruction Scope
SKILL.md and the included scripts keep scope to creating captcha-solving tasks and polling results from the CapSolver API. The instructions tell the agent to run scripts/solver.py with task-specific args (images, websiteURL, proxy, etc.). The skill does not instruct the agent to read unrelated system files or other environment variables beyond API_KEY, nor to send data to unexpected endpoints (only api.capsolver.com).
✓ Install Mechanism
There is no download/install step in the manifest; the package contains a Python script and a requirements.txt (requests, python-dotenv). No external or obscure URLs are downloaded at install time. This is lower risk from an install-mechanism perspective.
⚠ Credentials
The runtime requires an API_KEY (and the code loads .env) which is proportionate to contacting the CapSolver service. But the registry metadata claims no required env vars while SKILL.md and the code require API_KEY — this mismatch is a red flag (manifest/instruction inconsistency). The tool also accepts sensitive values as command-line args (proxies, awsKey/awsIv/awsApiKey, awsExistingToken) which are plausible for certain captcha types but mean you must not pass unrelated secrets inadvertently.
✓ Persistence & Privilege
The skill does not request 'always: true' and is user-invocable; it can be invoked autonomously (default), which is normal for skills. Combined with network access and an external API key, autonomous invocation increases blast radius, but this is expected for a networked service and not by itself a disqualifying privilege.
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 capsolver-captcha-solver
  3. 安装完成后，直接呼叫该 Skill 的名称或使用 /capsolver-captcha-solver 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
v1.0.1
- Initial release with CapSolver integration for automated captcha solving. - Added requirements.txt for dependency management. - Added scripts/solver.py providing various command-line tools to solve different CAPTCHA types using the CapSolver API.
v1.0.0
Initial release of capsolver-captcha-solver. - Supports automatic solving of various captchas including Geetest, reCAPTCHA v2/v3, MTCaptcha, DataDome, AWS WAF, Cloudflare Turnstile, and Cloudflare Challenge. - Provides command-line usage examples for each supported CAPTCHA type. - Environment variable `API_KEY` required for authentication. - Includes multiple solving modes: image recognition and token retrieval. - Documentation covers specific flags and parameters for customizing solving behavior.
元数据
Slug capsolver-captcha-solver
版本 1.0.1
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 2
常见问题

CapSolver 是什么？

Use capsolver to automatically resolve Geetest, reCAPTCHA v2, reCAPTCHA v3, MTCaptcha, DataDome, AWS WAF, Cloudflare Turnstile, and Cloudflare Challenge, etc. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 309 次。

如何安装 CapSolver？

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

CapSolver 是免费的吗？

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

CapSolver 支持哪些平台？

CapSolver 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 CapSolver？

由 capsolver（@capsolver）开发并维护，当前版本 v1.0.1。

推荐 Skills
self-improving agent
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...
⬇ 463283 · by pskoett
Skill Vetter
Security-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope, and suspicious patterns.
⬇ 259410 · by spclaudehome
Polymarket
Query Polymarket prediction markets. Check odds, find trending markets, search events, track price movements.
⬇ 232503 · by joelchance
Self-Improving + Proactive Agent
Self-reflection + Self-criticism + Self-learning + Self-organizing memory. Agent evaluates its own work, catches mistakes, and improves permanently. Use when...
⬇ 200423 · by ivangdavila
Github
Interact with GitHub using the `gh` CLI. Use `gh issue`, `gh pr`, `gh run`, and `gh api` for issues, PRs, CI runs, and advanced queries.
⬇ 191113 · by steipete
ontology
Typed knowledge graph for structured agent memory and composable skills. Use when creating/querying entities (Person, Project, Task, Event, Document), linkin...
⬇ 190156 · by oswalpalash

💬 留言讨论