de en es ja ko ru zh

solvegate-mcp

Name: solvegate-mcp
Author: SolveGate

Official

by SolveGate

Overview Schema Source Related Servers Discussions Admin

JavaScript

Remote

solvegate-mcp

This is an MCP server that allows an agent to handle Cloudflare Turnstile: check whether a page has one, and if it does, clear it.

There are three tools. One of them requires no key.

Tool	What it does	Key required
`inspect_page`	Fetches a URL and reports whether Cloudflare Turnstile exists, its sitekey, how it renders, its `data-*` configuration, and whether the page is a full-page WAF challenge rather than a widget.	Not required
`solve_turnstile`	Clears a Turnstile or Turnstile WAF challenge and returns a token.	Required
`get_solve`	Retrieves a solve by ID — free, no additional billing.	Required

Installation

npx solvegate-mcp

For Claude Desktop / any MCP client, add the following to your configuration:

{
  "mcpServers": {
    "solvegate": {
      "command": "npx",
      "args": ["-y", "solvegate-mcp"],
      "env": { "SOLVEGATE_API_KEY": "sk_test_…" }
    }
  }
}

SOLVEGATE_API_KEY will still work even if inspect_page is not set. This is intentional: you should be able to see what this server does before deciding whether to sign up for anything.

You can get a key at solvegate.io. The free sk_test_ sandbox key requires no card and returns a deterministic token immediately. It is suitable for setting this up, and it never actually clears a real gate.

Related MCP server: cloudscraper-mcp

Why `inspect_page` exists

The most common mistake when using a solver is to call it before first verifying what is actually on the page. Most cases like this fall into three categories:

The page is protected by reCAPTCHA or hCaptcha, not Turnstile. These are different products from different vendors. This server does not solve them, and neither does SolveGate. inspect_page will clearly tell you which one it is instead of having you discover it through a 422 error.
It is a WAF challenge, not a widget. A full-page Cloudflare interstitial has no data-sitekey because the challenge belongs to a zone rule rather than an embedded widget. There is nothing to read from the page, and the gate type is waf, not turnstile.
The sitekey is rendered from a script after page load. In that case, it is not present in the served HTML at all. The tool will tell you that instead of guessing.

Inspecting a page is free, takes only a second, and is very often the entire solution.

What information it will not provide

Widget mode. The managed, non-interactive, and invisible states are stored on the widget within the Cloudflare account that owns the sitekey. The page only carries the key, so no tool can read the mode from the source alone — including this one or any other that claims to do so.

This does not affect solving: all three modes result in the same request. You provide a sitekey, a page URL, and receive a token back.

The one exception is the five documented dummy sitekeys from Cloudflare, whose widget types are published by Cloudflare. inspect_page reports these, and if you find one on a page, it indicates that you are looking at a staging build rather than production.

Configuration

Variable	Default	Description
`SOLVEGATE_API_KEY`	—	Your key. Use `sk_test_` for the sandbox and `sk_live_` for real gates.
`SOLVEGATE_BASE_URL`	`https://api.solvegate.io`	API base URL.
`SOLVEGATE_SITE_URL`	`https://solvegate.io`	Where `inspect_page` runs.
`SOLVEGATE_TIMEOUT_MS`	`30000`	Per-request timeout.
`SOLVEGATE_MAX_WAIT_MS`	`60000`	How long `solve_solveturnstile` waits for a pending solve.

The key is read from the environment and is never exposed as a tool argument. A tool argument is visible to the model and may appear in transcripts.

Behavior Notes

Solving is performed asynchronously internally. solve_turnstile creates the solve with async: true and polls the result, rather than holding a connection open across the entire attempt. An agent’s tool call has a timeout that it cannot control, and a synchronous solve can make a slow gate look like a dead tool. Additionally, if a timeout occurs, a solve reference remains available: the error will mention it, and get_solve can retrieve the solve rather than losing one that has already been billed.

Errors are returned as results, not exceptions. Every failure is returned as an isError result that includes SolveGate’s own error.code — such as balance_empty, unknown_sitekey, rate_limited, or forbidden_target. A model can act on these values, and there is no reliable action to take from a stack trace.

Tokens are single-use and short-lived. expiresAt is specified in Unix seconds. Once past this time, Cloudflare no longer accepts the token, and the solve record is just a history entry. Request a token at the moment you submit it.

Check mode before trusting a token. An sk_test_ key causes mode: "sandbox" to be returned, and a sandbox token will not clear a real gate. This field ensures that a test key used in a production configuration fails clearly and loudly, rather than silently.

Acceptable Use

Only solve against properties you own or are authorized to test. This server is meant for end-to-end tests that keep failing on a challenge, CI environments for your own staging, and synthetic monitoring of your own production. The acceptable use policy is the binding version. Targets on a denylist always return 403 forbidden_target, no matter what an agent requests.

Development

npm install
npm run build
npm test          # spawns the built server and speaks MCP to it
OFFLINE=1 npm test  # skips the two tests that reach the network

The tests run through a real stdio transport rather than the callback functions directly. This is intentional: a schema that the SDK rejects at registration, or a stray write to stdout — which corrupts the transport because stdout is the transport — only appears when something actually goes through the protocol.

Available Tools

3 tools

get_solveRetrieve a solve by idA

Read-only

Fetch a solve created earlier. Free — reading never re-bills. Use this when solve_turnstile timed out while still pending: the solve may have finished after the tool gave up, and the id is in that error message. A 404 means no such solve for this API key's workspace.

Name	Required	Description	Default
`id`	Yes	The solve id, e.g. slv_8Kd2aF9

Output Schema

Name	Required	Description
`id`	Yes
`mode`	Yes
`token`	Yes
`billed`	Yes
`status`	Yes
`summary`	Yes
`errorCode`	Yes
`expiresAt`	Yes

TDQS

A4.7/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already mark it read-only and open-world, but the description adds valuable behavioral context: it is free and never re-bills, and a 404 is scoped to the current API key's workspace. This goes beyond structured metadata without contradicting it.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Three concise sentences with no filler. The main purpose is front-loaded, followed by a concrete use case and an error interpretation; every sentence earns its place.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a single-parameter retrieval tool with an output schema and safety annotations, the description fully covers when to call it, what the id means, cost implications, and error semantics. Nothing essential is missing for correct invocation.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100% and the only parameter, id, is already clearly described with a format example. The description adds the useful hint that the id may come from a timeout error, but does not need to elaborate further since the schema carries the parameter meaning.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource ('Fetch a solve... by id') and distinguishes itself from siblings by describing exactly when it is the right tool. The mention of solve_turnstile and 404 behavior makes its role unambiguous.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Explicitly tells the agent when to use this tool: after solve_turnstile times out while the solve is pending, with the id coming from that error message. It also interprets a 404 response, giving clear operational context beyond the schema.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

inspect_pageInspect a page for Cloudflare TurnstileA

Read-only

Fetch a URL and report what its HTML says about Cloudflare Turnstile: whether a widget is present, its sitekey, whether it renders implicitly or via turnstile.render(), the widget's data-* configuration, whether the sitekey is one of Cloudflare's documented test keys, and whether the page is a full-page WAF challenge rather than a widget. Free, no API key required. Use this before solving when the sitekey is unknown. It CANNOT determine the widget's mode (managed / non-interactive / invisible) — that is stored in the Cloudflare account that owns the sitekey and is not in the page at all.

Name	Required	Description	Default
`url`	Yes	The page URL to inspect, e.g. https://example.com/login

Output Schema

Name	Required	Description
`present`	Yes	Turnstile was found on the page
`summary`	Yes	One line a human can read
`finalUrl`	Yes
`sitekeys`	Yes	Sitekeys found, in document order
`rendering`	Yes
`httpStatus`	Yes
`otherCaptcha`	Yes	Other CAPTCHA products on the page — we do not solve these
`wafChallenge`	Yes	The response is a Cloudflare WAF challenge interstitial
`responseFieldPresent`	Yes

TDQS

A4.7/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Beyond the readOnlyHint, it adds meaningful behavioral specifics: it performs an HTML inspection rather than solving a challenge, requires no API key, is free, and has a specific hard limitation around widget mode. These details substantially describe side effects and boundaries beyond annotations.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Although the description is long, each clause delivers a distinct, useful fact: what is fetched, what is reported, cost/auth requirements, usage timing, and a critical limitation. The most important action and resource are front-loaded, and no sentence is slack.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given a single required parameter, an output schema, and annotations, the description covers everything an agent needs: when to call it, what it returns, its constraints, and what it cannot infer. The mitigation of return format details is appropriate.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The schema fully describes the single 'url' parameter with an example, so parameter documentation coverage is 100%. The description does not add much beyond the schema, but the schema already carries the necessary semantic weight. This aligns with the baseline for high schema coverage.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific verb ('Fetch') and a concrete result ('report what its HTML says about Cloudflare Turnstile') and enumerates exact outputs: widget presence, sitekey, render method, data-* configuration, test keys, and full-page WAF challenge. This clearly separates it from sibling solving/result-retrieval tools.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It explicitly says 'Use this before solving when the sitekey is unknown,' giving a direct invocation condition. It also provides an exclusion: it cannot determine the widget's mode because that data is not in the page, so agents know when this tool is not sufficient.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

solve_turnstileSolve a Cloudflare Turnstile or WAF challengeA

Clear a Cloudflare challenge and return a token. gate is "turnstile" for an embedded widget (there is a data-sitekey in the page) or "waf" for a full-page Cloudflare interstitial. sitekey and url are both required; the url must be the page the challenge actually renders on, because a sitekey is bound to its configured hostnames. Tokens are single-use and expire in minutes — request one at the moment you submit. Failed solves are never billed. Requires SOLVEGATE_API_KEY.

Name	Required	Description
`url`	Yes	The page URL the challenge renders on
`gate`	Yes	Which challenge type to clear
`proxy`	No	Optional egress proxy URL to solve through
`action`	No	Optional Turnstile action label, if the widget sets one
`sitekey`	Yes	The target's sitekey, e.g. 0x4AAAAAAA…

Output Schema

Name	Required	Description
`id`	Yes
`mode`	Yes	sandbox tokens NEVER clear a real gate
`token`	Yes	Submit this as the cf-turnstile-response field
`billed`	Yes
`status`	Yes
`solveMs`	Yes
`summary`	Yes
`errorCode`	Yes
`expiresAt`	Yes	Unix seconds after which the token is worthless

TDQS

A4.7/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations indicate non-read-only and non-idempotent behavior, and the description adds meaningful operational context: tokens are single-use and expire in minutes, failed solves are never billed, SOLVEGATE_API_KEY is required, and sitekeys are bound to hostnames. This goes well beyond the annotations and clarifies important side effects and constraints.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Four sentences with no filler: purpose is front-loaded, then parameter guidance, then token lifecycle and error/cost behavior. Every sentence adds useful information and the structure is easy to parse.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

With an output schema and useful annotations present, the description supplies the critical missing context: token expiry timing, no-billing on failed solves, API key requirement, and hostname binding. Nothing an agent needs to invoke this tool correctly appears to be missing.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3. The description adds extra semantic value by explaining the `gate` enum values, reinforcing that `sitekey` and `url` are both required, and explaining why `url` must be the actual rendering page. Some parameters like `proxy` and `action` receive no additional description, but the schema already documents them adequately.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description opens with a specific verb-resource pair ('Clear a Cloudflare challenge and return a token') and then defines the two `gate` modes, clearly distinguishing an embedded widget from a full-page interstitial. This makes the tool's purpose immediately understandable and separates it from generic 'solve' tools.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description gives actionable selection guidance: it explains when to use 'turnstile' versus 'waf', and instructs the agent to request tokens at submission time due to expiry. It does not explicitly name sibling tools as alternatives or state when not to use this tool, so it stops short of full when/when-not guidance.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

2 tool updatesv0.1.2Aug 29, 2026

Changedget_solve6 fields changed

removedOutput schema / properties / errorCode / anyOf

Removed value: -[
-  {
-    "type": "string"
-  },
-  {
-    "type": "null"
-  }
-]

addedOutput schema / properties / errorCode / type
```
Added value: +[
+  "string",
+  "null"
+]
```

removedOutput schema / properties / expiresAt / anyOf

Removed value: -[
-  {
-    "type": "number"
-  },
-  {
-    "type": "null"
-  }
-]

addedOutput schema / properties / expiresAt / type
```
Added value: +[
+  "number",
+  "null"
+]
```

removedOutput schema / properties / token / anyOf

Removed value: -[
-  {
-    "type": "string"
-  },
-  {
-    "type": "null"
-  }
-]

addedOutput schema / properties / token / type
```
Added value: +[
+  "string",
+  "null"
+]
```

Changedsolve_turnstile8 fields changed

removedOutput schema / properties / errorCode / anyOf

Removed value: -[
-  {
-    "type": "string"
-  },
-  {
-    "type": "null"
-  }
-]

addedOutput schema / properties / errorCode / type
```
Added value: +[
+  "string",
+  "null"
+]
```

removedOutput schema / properties / expiresAt / anyOf

Removed value: -[
-  {
-    "type": "number"
-  },
-  {
-    "type": "null"
-  }
-]

addedOutput schema / properties / expiresAt / type
```
Added value: +[
+  "number",
+  "null"
+]
```

removedOutput schema / properties / solveMs / anyOf

Removed value: -[
-  {
-    "type": "number"
-  },
-  {
-    "type": "null"
-  }
-]

addedOutput schema / properties / solveMs / type
```
Added value: +[
+  "number",
+  "null"
+]
```

removedOutput schema / properties / token / anyOf

Removed value: -[
-  {
-    "type": "string"
-  },
-  {
-    "type": "null"
-  }
-]

addedOutput schema / properties / token / type
```
Added value: +[
+  "string",
+  "null"
+]
```

3 tool updatesv0.1.1Aug 25, 2026
- First observedget_solve
- First observedinspect_page
- First observedsolve_turnstile

Deploy Server

Give agents eyes on the webHosted screenshot MCP. Send a URL, get a pixel-perfect image. $0.002 per grab. Works with Claude, Cursor, and Codex.grabbit.liveAd

TDQS

A4.8/5.0

Scored 2026-08-31 10:15 across 3 tools

Disambiguation5/5

Each tool serves a distinct purpose: inspect_page reads page HTML, solve_turnstile solves the challenge, and get_solve retrieves a previous result. There is no functional overlap between them, and their descriptions clearly differentiate when to use each.

Naming Consistency5/5

All tool names follow the verb_noun pattern in snake_case: inspect_page, solve_turnstile, get_solve. The pattern is consistent and predictable, making it easy to infer tool behavior from the name.

Tool Count5/5

With only 3 tools, the server remains tightly focused on the Turnstile solving workflow. Each tool is essential and there is no bloat or redundancy, so the count is perfectly matched to the scope.

Completeness5/5

The tool surface covers the full lifecycle: pre-flight inspection (inspect_page), actual solving (solve_turnstile), and post-hoc retrieval (get_solve). This is a complete workflow with no obvious dead ends or missing operations for the stated purpose.

Maintenance

ActivityStale

ResponsivenessNo issues

Resources

Related MCP Connectors

cloudflare-solver B
com.jojapi
Cloudflare Solver: Scraping API designed to bypass Cloudflare protection.
Web Scraping Browser Automation
11
https://mcp.jojapi.com/cloudflare-solver
Captcha Solver B
io.github.mcp-dir
Solve reCAPTCHA v2/v3, hCaptcha, Cloudflare Turnstile and image captchas. Platform-hosted, no creden
Browser Automation Web Scraping
121
https://api.mcp.ai/p_captcha
Solvecha A
com.solvecha
Solve reCAPTCHA image grids, text captchas, GeeTest and FunCAPTCHA for AI agents. Free tier.
Browser Automation Web Scraping Penetration Testing
61
https://api.solvecha.com/mcp
Human Handoff relay A
com.botrecipe
Hands CAPTCHAs, 2FA codes and approvals from your agent to your human with one tap.
Agent Orchestration Autonomous Agents Security
51
https://relay.cloudforbots.com/mcp

Related MCP Servers

sessemi-mcpofficial
Web Scraping Browser Automation
sessemi
A
license
A
quality
D
maintenance
Enables AI agents to scrape websites protected by Cloudflare, DataDome, and Akamai, bypassing anti-bot measures to retrieve page content.
Updated 5 months ago (2026-05-22 09:45 UTC)
2
MIT
cloudscraper-mcp
Browser Automation Web Scraping
LLMTooling
F
license
Not graded
quality
D
maintenance
Enables AI agents to bypass Cloudflare protection and scrape web content, returning clean Markdown with smart chunking and file export.
Updated 5 months ago (2026-05-21 00:25 UTC)
1
-
disasm.dev MCP serverofficial
Browser Automation Web Scraping
disasm-dev
A
license
A
quality
D
maintenance
Enables AI assistants to generate DataDome and Incapsula clearance tokens without a headless browser, for development and prototyping.
Updated 4 months ago (2026-06-25 13:11 UTC)
6
MIT
webshot-mcp
Browser Automation Web Scraping
Brandon-35
A
license
Not graded
quality
C
maintenance
Enables agents to take screenshots and extract text from web pages using a persistent Chrome browser via CDP, with built-in SSRF protection.
Updated 2 months ago (2026-08-18 04:18 UTC)
9 npm
MIT

solvegate-mcp

Installation

Why inspect_page exists

What information it will not provide

Configuration

Behavior Notes

Acceptable Use

Development

Links

Available Tools

Output Schema

TDQS

Output Schema

TDQS

Output Schema

TDQS

Tool Schema Changelog

TDQS

Maintenance

Resources

Related MCP Connectors

Related MCP Servers

sessemi-mcpofficial

cloudscraper-mcp

disasm.dev MCP serverofficial

webshot-mcp

Why `inspect_page` exists