JEV

JEV Evaluate API MCP

Evaluate text against predefined criteria or your own questions. Call the JEV Content Evaluation API via REST or MCP with your Nexscope API key.

Cost10 credits / callOnline Test

JEV Evaluate API AI Prompts & Use Cases

After connecting this MCP tool to your AI assistant, you can ask the questions below. Replace bracketed placeholders with your own inputs and use the supported values in Arguments.

New to Nexscope MCP? Follow the connection guide, then use the prompts below. The Integration and Parameters & Response tabs contain the connection steps, call example, and technical reference.

1. Check A Draft Response

Prompt: Check [draft response] against the response-safety preset. Explain the returned decision without sending or publishing the draft.

2. Triage Supplied IP Evidence

Prompt: Evaluate [supplied product and IP evidence] with the IP-risk preset. Keep uncertainty explicit; do not turn the result into a legal clearance.

3. Evaluate SEO Relevance

Prompt: Evaluate [page content and search intent] with the SEO-relevance preset, reporting only supported dimensions.

4. Check A Claim Against Supplied Evidence

Prompt: Evaluate [claim or quotation] against [supplied source evidence]. Report missing evidence instead of filling it from memory.

5. Check Keyword Business Relevance

Prompt: Evaluate [candidate keywords with IDs] against [business facts]. Keep decisions tied to the supplied IDs and facts.

6. Check Whether Keywords Belong Together

Prompt: Evaluate [candidate keywords and supplied SERP evidence] for same-page clustering. Explain the returned fit decisions without inventing search results.

7. Check Product And Keyword Fit

Prompt: Evaluate [candidate keywords with IDs] against [Amazon product facts]. Keep any fit decision limited to the supplied product evidence.

8. Evaluate Evidence With Yes-Or-No Questions

Prompt: Evaluate [shared evidence] using [one to 32 questions] that request yes-or-no assessments. Give each question a short identifier and a specific question; include my definitions of yes and no when supplied. Use custom evaluation without a preset, and report the returned answers without executing an action.

9. Ask Custom Choice Questions

Prompt: Evaluate [shared evidence] using [one to 32 questions] that ask you to choose among my supplied options. Give each question an identifier and two to 255 labeled options, with descriptions where needed. Use custom evaluation without a preset and report only the returned choices.

10. Ask Custom Score Questions

Prompt: Evaluate [shared evidence] using [one to 32 questions] on my supplied scoring scales. Give each question an identifier and two to 32 ordered level descriptions, starting at zero. Explain that the returned score is a probability-weighted average, not a measured fact. Use custom evaluation without a preset.

What You Need to Provide

Choose one mode: a preset with its matching input, or shared context with typed custom questions. Never mix the two modes.

Preset inputs and custom question structures must follow the nested schema. Evaluation is advisory and does not execute another API or publish content.

Step 1 · Connect

Connect Your AI Assistant

Connect a compatible MCP client to Nexscope, then select this tool from your assistant. Use the connection settings in your client to complete authentication.

POST https://api.nexscope.ai/mcp
MCP-Protocol-Version: 2025-11-25
Content-Type: application/json

Authentication

Use OAuth 2.1 Authorization Code with PKCE and the mcp:tools scope. API keys remain supported as a compatibility option.

Authorization: Bearer nk-xxxxxxxxxxxxxxxxx

Using an API key? Get your key from API Access and include it in the Authorization header shown above.

Call this MCP tool

MCP tool name

nexscope_jev_evaluate

MCP endpoint

https://api.nexscope.ai/mcp

JSON-RPC method

tools/call

Initialize an MCP 2025-11-25 Streamable HTTP session at https://api.nexscope.ai/mcp, then use tools/call.

{
  "jsonrpc": "2.0",
  "id": "jev-evaluate-1",
  "method": "tools/call",
  "params": {
    "name": "nexscope_jev_evaluate",
    "arguments": {
      "input": {
        "content": "Never share passwords. Contact customer support through the official website."
      },
      "useCase": "response_guardrails"
    }
  }
}

MCP 2025-11-25 · Streamable HTTP

Progress is delivered on the original POST SSE response. GET SSE carries notifications; reconnecting must query existing asynchronous work rather than submit again.

Arguments

Name	Type	Required	Description
`input`	`object`	optional	Fields depend on the selected preset; see the mode schemas below.
`questions`	`object`	optional	Required in custom mode. 1–32 independent questions, keyed by caller-defined IDs.
`state`	`string \| object \| array`	optional	Required in custom mode. Non-empty context shared by every question; supplied data is sent to TypeSafe. Minimum length: 1 Minimum items: 1
`useCase`	`string`	optional	Required with input in preset mode. One of seven server-owned presets. Allowed: response_guardrails, ip_risk_triage, seo_relevance, evidence_check, keyword_business_relevance, keyword_cluster_fit, amazon_keyword_product_fit

Preset mode

response_guardrails

Name	Type	Required	Description
`input`	`object`	optional	Checks credential_exposure, harmful_instruction, deceptive_claim, instruction_override, and severity. The fixed server policy maps the checks to pass, review, or block.
`input.content`	`string`	required	Authorized response or draft content. Minimum length: 1 Maximum length: 6000
`useCase`	`string`	optional	Required with input in preset mode. One of seven server-owned presets. Constant: response_guardrails

ip_risk_triage

Name	Type	Required	Description
`input`	`object`	optional	Checks trademark, copyright, patent, other IP, and insufficient-context signals plus risk type and severity. The result always requires human review and is not legal clearance.
`input.text`	`string`	required	Authorized product or description text for advisory IP-risk triage. Minimum length: 1 Maximum length: 4000
`useCase`	`string`	optional	Required with input in preset mode. One of seven server-owned presets. Constant: ip_risk_triage

seo_relevance

Name	Type	Required	Description
`input`	`object`	optional	Requires query and at least one supplied title, metaDescription, or contentExcerpt. Scores relevance only; it does not predict rankings, traffic, search volume, or intent data.
`input.contentExcerpt`	`string`	optional	Content excerpt value used by this API operation. Minimum length: 1 Maximum length: 4000
`input.metaDescription`	`string`	optional	Meta description value used by this API operation. Minimum length: 1 Maximum length: 1000
`input.query`	`string`	required	Query value used by this API operation. Minimum length: 1 Maximum length: 500
`input.title`	`string`	optional	Title value used by this API operation. Minimum length: 1 Maximum length: 300
`useCase`	`string`	optional	Required with input in preset mode. One of seven server-owned presets. Constant: seo_relevance

Any of option 1 — requires title

Name	Type	Required	Description
`input`	`any`	optional	Required with useCase in preset mode. Strict preset-specific input. Keyword presets accept only bounded keyword, candidate, SERP-evidence, and product-fact fields.

Any of option 2 — requires metaDescription

Name	Type	Required	Description
`input`	`any`	optional	Required with useCase in preset mode. Strict preset-specific input. Keyword presets accept only bounded keyword, candidate, SERP-evidence, and product-fact fields.

Any of option 3 — requires contentExcerpt

Name	Type	Required	Description
`input`	`any`	optional	Required with useCase in preset mode. Strict preset-specific input. Keyword presets accept only bounded keyword, candidate, SERP-evidence, and product-fact fields.

evidence_check

Name	Type	Required	Description
`input`	`object`	optional	Assesses the claim only against sourceText. If supplied quote is absent after whitespace normalization, the server returns deterministic quote_not_found without a model call.
`input.claim`	`string`	required	Claim value used by this API operation. Minimum length: 1 Maximum length: 1000
`input.quote`	`string`	optional	Optional quote that must occur in sourceText after whitespace normalization. Minimum length: 1 Maximum length: 1000
`input.sourceText`	`string`	required	Source text value used by this API operation. Minimum length: 1 Maximum length: 6000
`useCase`	`string`	optional	Required with input in preset mode. One of seven server-owned presets. Constant: evidence_check

keyword_business_relevance

Name	Type	Required	Description
`input`	`object`	optional	Required with useCase in preset mode. Strict preset-specific input. Keyword presets accept only bounded keyword, candidate, SERP-evidence, and product-fact fields.
`input.businessDescription`	`string`	required	Business description value used by this API operation. Minimum length: 1 Maximum length: 4000
`input.exclusions`	`array`	required	Exclusions value used by this API operation. Maximum items: 50
`input.exclusions[]`	`string`	required	Item in input exclusions. Minimum length: 1 Maximum length: 200
`input.keywords`	`array`	required	Keywords value used by this API operation. Minimum items: 1 Maximum items: 10
`input.keywords[]`	`object`	required	Item in input keywords.
`input.keywords[].id`	`string`	required	Id value used by this API operation. Pattern: ^[A-Za-z0-9_-]{1,64}$
`input.keywords[].keyword`	`string`	required	Keyword value used by this API operation. Minimum length: 1 Maximum length: 500
`input.serpEvidence`	`array`	optional	Serp evidence value used by this API operation. Maximum items: 30
`input.serpEvidence[]`	`object`	required	Item in input serpEvidence.
`input.serpEvidence[].id`	`string`	required	Id value used by this API operation. Pattern: ^[A-Za-z0-9_-]{1,64}$
`input.serpEvidence[].keywordId`	`string`	required	Keyword id value used by this API operation. Pattern: ^[A-Za-z0-9_-]{1,64}$
`input.serpEvidence[].source`	`string`	required	Source value used by this API operation. Minimum length: 1 Maximum length: 200
`input.serpEvidence[].text`	`string`	required	Text value used by this API operation. Minimum length: 1 Maximum length: 2000
`useCase`	`string`	optional	Required with input in preset mode. One of seven server-owned presets. Constant: keyword_business_relevance

keyword_cluster_fit

Name	Type	Required	Description
`input`	`object`	optional	Required with useCase in preset mode. Strict preset-specific input. Keyword presets accept only bounded keyword, candidate, SERP-evidence, and product-fact fields.
`input.candidates`	`array`	required	Candidates value used by this API operation. Minimum items: 1 Maximum items: 5
`input.candidates[]`	`object`	required	Item in input candidates.
`input.candidates[].id`	`string`	required	Id value used by this API operation. Pattern: ^[A-Za-z0-9_-]{1,64}$
`input.candidates[].keyword`	`string`	required	Keyword value used by this API operation. Minimum length: 1 Maximum length: 500
`input.keywords`	`array`	required	Keywords value used by this API operation. Minimum items: 1 Maximum items: 10
`input.keywords[]`	`object`	required	Item in input keywords.
`input.keywords[].id`	`string`	required	Id value used by this API operation. Pattern: ^[A-Za-z0-9_-]{1,64}$
`input.keywords[].keyword`	`string`	required	Keyword value used by this API operation. Minimum length: 1 Maximum length: 500
`useCase`	`string`	optional	Required with input in preset mode. One of seven server-owned presets. Constant: keyword_cluster_fit

amazon_keyword_product_fit

Name	Type	Required	Description
`input`	`object`	optional	Required with useCase in preset mode. Strict preset-specific input. Keyword presets accept only bounded keyword, candidate, SERP-evidence, and product-fact fields.
`input.keywords`	`array`	required	Keywords value used by this API operation. Minimum items: 1 Maximum items: 10
`input.keywords[]`	`object`	required	Item in input keywords.
`input.keywords[].id`	`string`	required	Id value used by this API operation. Pattern: ^[A-Za-z0-9_-]{1,64}$
`input.keywords[].keyword`	`string`	required	Keyword value used by this API operation. Minimum length: 1 Maximum length: 500
`input.marketplace`	`string`	required	Marketplace value used by this API operation. Allowed: US, UK, DE, FR, JP, CA, IT, ES, MX, IN
`input.productFacts`	`array`	required	Product facts value used by this API operation. Minimum items: 1 Maximum items: 30
`input.productFacts[]`	`object`	required	Item in input productFacts.
`input.productFacts[].id`	`string`	required	Id value used by this API operation. Pattern: ^[A-Za-z0-9_-]{1,64}$
`input.productFacts[].source`	`string`	required	Source value used by this API operation. Allowed: title, bullet, attribute, description
`input.productFacts[].value`	`string`	required	Value value used by this API operation. Minimum length: 1 Maximum length: 2000
`useCase`	`string`	optional	Required with input in preset mode. One of seven server-owned presets. Constant: amazon_keyword_product_fit

Custom mode

Response fields

Treat structuredContent as the programmatic API response payload.

Name	Type	Required	Description
`code`	`integer`	required	Zero on success; non-zero for validation, access, quota, or availability errors.
`data`	`object`	required	Fixed server policies cover response safety, IP triage, SEO relevance, evidence checks, business-keyword relevance, keyword same-page clustering, and Amazon product-keyword fit. Keyword presets return bounded rows keyed only by caller-supplied IDs.
`data.checks`	`object`	required	Empty in custom mode; answers are in result.answers. Fixed Noul, Score, or Choice checks for the preset: response risks and severity; IP signals, risk_type, severity, and insufficient_context; supplied SEO dimensions; or evidence verdict. Deterministic quote_not_found returns an empty object.
`data.decision`	`string`	required	Presets return pass/review/block. Custom mode returns evaluated, which means answers are available, not that an action is approved. Allowed: pass, review, block, evaluated
`data.model`	`string \| null`	required	Resolved model version, or null for deterministic quote_not_found.
`data.policyVersion`	`string`	required	jev-evaluation-v1 for original presets; jev-keyword-tools-v1 for keyword presets; jev-custom-v1 for custom questions.
`data.reasonCodes`	`array`	required	Fixed reason codes without submitted text.
`data.reasonCodes[]`	`string`	required	Fixed reason codes without submitted text.
`data.result`	`object`	required	Preset-specific bounded result. Keyword presets return result.rows with caller-supplied IDs, fixed labels, confidence, and fixed reason or evidence references.
`data.result.answers`	`object`	optional	Custom mode only. Answers keyed by question ID: noul returns {type, noul}; choice returns {type, choice, confidence, probabilities}; score returns {type, score, confidence, probabilities, legend}. No automatic policy or token usage is returned.
`data.result.rows`	`array`	optional	Rows value used by this API operation. Maximum items: 10
`data.result.rows[]`	`any`	required	Item in data result rows.
`data.useCase`	`string`	required	Use case value used by this API operation. Allowed: response_guardrails, ip_risk_triage, seo_relevance, evidence_check, keyword_business_relevance, keyword_cluster_fit, amazon_keyword_product_fit, custom
`msg`	`string`	required	Response message.

Business relevance row

Name	Type	Required	Description
`data.result.rows[]`	`object`	required	Item in data result rows.
`data.result.rows[].confidence`	`number`	required	Confidence value used by this API operation. Minimum: 0 Maximum: 1
`data.result.rows[].evidenceRefs`	`array`	required	Evidence refs value used by this API operation.
`data.result.rows[].evidenceRefs[]`	`string`	required	Item in data result rows[] evidenceRefs.
`data.result.rows[].id`	`string`	required	Id value used by this API operation. Pattern: ^[A-Za-z0-9_-]{1,64}$
`data.result.rows[].label`	`string`	required	Label value used by this API operation. Allowed: relevant, irrelevant, uncertain
`data.result.rows[].reasonCodes`	`array`	required	Reason codes value used by this API operation.
`data.result.rows[].reasonCodes[]`	`string`	required	Item in data result rows[] reasonCodes.

Keyword cluster fit row

Name	Type	Required	Description
`data.result.rows[]`	`object`	required	Item in data result rows.
`data.result.rows[].candidateId`	`string \| null`	required	Candidate id value used by this API operation.
`data.result.rows[].confidence`	`number`	required	Confidence value used by this API operation. Minimum: 0 Maximum: 1
`data.result.rows[].id`	`string`	required	Id value used by this API operation. Pattern: ^[A-Za-z0-9_-]{1,64}$
`data.result.rows[].label`	`string`	required	Label value used by this API operation. Allowed: same_page, separate_page, uncertain
`data.result.rows[].reasonCodes`	`array`	required	Reason codes value used by this API operation.
`data.result.rows[].reasonCodes[]`	`string`	required	Item in data result rows[] reasonCodes.
`data.result.rows[].samePage`	`boolean`	required	Same page value used by this API operation.
`data.result.rows[].sameTopic`	`boolean`	required	Same topic value used by this API operation.

Amazon keyword product fit row

Name	Type	Required	Description
`data.result.rows[]`	`object`	required	Item in data result rows.
`data.result.rows[].confidence`	`number`	required	Confidence value used by this API operation. Minimum: 0 Maximum: 1
`data.result.rows[].evidenceRefs`	`array`	required	Evidence refs value used by this API operation.
`data.result.rows[].evidenceRefs[]`	`string`	required	Item in data result rows[] evidenceRefs.
`data.result.rows[].id`	`string`	required	Id value used by this API operation. Pattern: ^[A-Za-z0-9_-]{1,64}$
`data.result.rows[].label`	`string`	required	Label value used by this API operation. Allowed: match, partial, conflict, unknown
`data.result.rows[].reasonCodes`	`array`	required	Reason codes value used by this API operation.
`data.result.rows[].reasonCodes[]`	`string`	required	Item in data result rows[] reasonCodes.

Format Illustration

Illustrative example with fictional data, not a live tool call.

Illustrative Question

Check the statement “Never share passwords. Contact customer support through the official website.” with the response-safety preset.

Illustrative Response Explanation

The illustrative result returns pass with the reason that no rule was triggered. This advisory decision does not send the text or approve another action.

About this sample: Illustrative format example with fictional values, not a live tool call. Replace example inputs with your own permitted assets or public identifiers. Model and catalog labels in this illustration are fictional, not identifiers to select for a live request.

Published sample data is not available for this page. A format illustration or workflow guide is not a substitute for a published result.

View the MCP Request & Response Format

Examples