一款强大的研究助手，可与 Cline 和 Claude Desktop 集成！利用 Perplexity AI 进行智能搜索、文档检索、API 发现和代码现代化协助——所有这些都在您编写代码时完成。
无缝上下文跟踪：在 SQLite 数据库中维护对话历史记录，以便在多个查询中提供一致的响应
高级查询处理：使用 Perplexity 的 Sonar 模型进行复杂的推理，并为复杂问题提供详细的答案
智能速率管理：通过指数退避实现自适应速率限制，以最大限度地提高 API 使用率，而不会达到限制
高性能网络：通过连接池和自动重试逻辑优化 API 调用，实现可靠运行
Related MCP server: Perplexity Tool for Claude Desktop
执行常规搜索查询，获取任何主题的全面信息。示例展示了如何使用不同的详细程度（简要、常规、详细）来获取定制的响应。
检索特定技术、库或 API 的文档和使用示例。该示例演示了如何获取 React Hooks 的全面文档，包括最佳实践和常见陷阱。
发现并评估可集成到项目中的 API。示例展示了如何查找支付处理 API，并详细分析了其功能、定价和集成复杂性。
分析代码中已弃用的功能或模式，并提供迁移指导。示例演示了如何检查 React 类组件和生命周期方法，以找到现代替代方案。
如果尚未安装，请先安装 Node.js（来自 nodejs.org）
克隆 repo
安装依赖项并构建
从https://www.perplexity.ai/settings/api获取 Perplexity API 密钥
在适合您的操作系统的位置创建 MCP 设置文件：
要与 Claude Desktop 一起使用，请添加服务器配置：
要与 Cline 一起使用，请添加到 mcpServers：
{
"mcpServers": {
"perplexity-server": {
"command": "node",
"args": ["[path/to/researcher-mcp/build/index.js]"],
"env": {
"PERPLEXITY_API_KEY": "pplx-...",
"PERPLEXITY_MODEL": "sonar-reasoning" // you can use different models
},
"disabled": false,
"alwaysAllow": [],
"autoApprove": [
"search",
"get_documentation",
"find_apis",
"check_deprecated_code",
"get_request_status"
]
}
}
}构建服务器：npm run build
Check if code or dependencies might be using deprecated features
| Name | Required | Description | Default |
|---|---|---|---|
| code | Yes | The code snippet or dependency to check | |
| technology | No | The technology or framework context (e.g., 'React', 'Node.js') |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description does not disclose behavioral traits beyond its basic function. It does not state if the check is read-only, whether it executes the code, what input format is expected, or what output is produced. With no annotations, this lack of detail is a significant gap.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, focused sentence with no redundant information. It gets straight to the point.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Despite the small parameter count, the description lacks crucial information about the tool's behavior and return value. Without an output schema or annotations, the agent is left without a clear picture of what to expect. The description is too minimal to be complete.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema fully describes both parameters (code and technology), so the description doesn't need to elaborate. The description adds no additional parameter context beyond the schema, aligning with the baseline score of 3.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's function: checking code or dependencies for deprecated features. It distinguishes from sibling tools like search and get_documentation by focusing on deprecation analysis rather than general information retrieval.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
No guidance is provided on when to use this tool versus alternatives like search or get_documentation. The description only states what it does, implying usage but offering no exclusions or comparison.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Find and evaluate APIs that could be integrated into a project
| Name | Required | Description | Default |
|---|---|---|---|
| requirement | Yes | The functionality or requirement you're looking to fulfill | |
| context | No | Additional context about the project or specific needs |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full burden of behavioral disclosure, but it only states the purpose. It doesn't mention whether the tool performs live searches, what evaluation criteria are used, what output format is returned, or any side effects. This is a significant lack of transparency.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, concise sentence that is front-loaded and directly conveys the tool's purpose. Every word earns its place, and there is no unnecessary detail or fluff.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the lack of annotations and output schema, the description should compensate by explaining return values, research process, or usage context. It does none of this, leaving the agent without a clear picture of the tool's output, behavior, or when to choose it over siblings.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The input schema already describes both parameters (requirement and context) with 100% coverage, so the baseline is 3. The tool description adds no additional meaning beyond the schema; it doesn't explain how the parameters influence behavior or results.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool finds and evaluates APIs for project integration, using a specific verb and resource. It distinguishes itself from sibling tools like general search and get_documentation, though it doesn't detail the evaluation criteria or depth.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description gives no guidance on when to use this tool versus alternatives. It doesn't mention any exclusions, prerequisites, or comparisons with sibling tools like search or get_documentation, leaving the appropriate usage context unclear.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get documentation and usage examples for a specific technology, library, or API
| Name | Required | Description | Default |
|---|---|---|---|
| query | Yes | The technology, library, or API to get documentation for | |
| context | No | Additional context or specific aspects to focus on |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden of behavioral disclosure. It only restates the primary function and reveals nothing about potential side effects, output format, source reliability, or limitations. The agent is left guessing about what 'documentation' entails (e.g., official docs, community examples, version specifics).
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, concise sentence that front-loads the main verb and resource. There is no wasted wording or unnecessary detail.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a simple tool with full schema coverage and no output schema, the description is minimally viable but lacks usage guidance and behavioral context. It could be more complete by explaining what types of documentation are returned or how it differs from sibling tools.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100% for both parameters, so the baseline is 3. The description adds no additional meaning beyond what the schema already provides; it merely reiterates 'specific technology, library, or API' which matches the 'query' parameter description.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the verb ('Get') and the resource ('documentation and usage examples') for a 'specific technology, library, or API.' This is specific and distinguishes it from generic search, though it doesn't explicitly call out sibling tools.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
No guidance is given on when to use this tool versus alternatives like search, find_apis, or chat_perplexity. The description implies a documentation-focused purpose but provides no exclusions, prerequisites, or contextual cues for selection.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Perform a general search query to get comprehensive information on any topic
| Name | Required | Description | Default |
|---|---|---|---|
| query | Yes | The search query or question | |
| detail_level | No | Optional: Desired level of detail (brief, normal, detailed) |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden of disclosing behavioral traits. It only promises 'comprehensive information' without noting whether the tool is read-only, what data source it accesses, whether it has rate limits, or what the response format looks like.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The one-sentence description is efficient and front-loaded with the verb and resource. It contains no filler, though its brevity limits the amount of actionable guidance.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the simple schema and lack of an output schema, the description still leaves important gaps: it does not explain what kind of results are returned, when this tool should be preferred over sibling tools, or how the optional detail_level affects behavior.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The input schema already documents both parameters with descriptions, giving 100% coverage. The description adds only generic context ('any topic') and does not enrich the meaning of the parameters beyond what the schema provides.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description states a clear action ('Perform a general search query') and resource ('search') to gather information. However, it does not differentiate this from sibling tools like get_documentation or find_apis, so it stops short of full clarity.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
No guidance is provided on when to choose this tool over alternatives. The phrase 'any topic' is extremely broad and gives no context or exclusions, leaving the agent without direction.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
check_deprecated_codefind_apisget_documentationsearchScored across 4 tools
Each tool has a clearly distinct purpose with no overlap: checking deprecated code, finding APIs, getting documentation, and performing general searches. The descriptions make it easy for an agent to select the right tool for each specific research task.
Three tools follow a consistent verb_noun pattern (check_deprecated_code, find_apis, get_documentation), but 'search' deviates as a single verb without an object. This minor inconsistency slightly affects predictability, though all names remain readable.
With only 4 tools, the set feels thin for a research server that aims to cover broad information-gathering tasks. While each tool is useful, the scope suggests more specialized research operations could be missing, making it borderline appropriate.
The tools cover key research functions like checking deprecations, finding APIs, getting docs, and general searches, but there are notable gaps. For example, missing tools for comparing technologies, validating information sources, or tracking research progress limit comprehensive workflow coverage.
Real-time web search, reasoning, and research through Perplexity's API
Persistent memory for Claude Code and Cursor. Stop re-explaining your project every session.
Persistent context for Claude. Your AI always knows your projects and next actions across sessions.
Enable AI assistants to perform web searches using Perplexity's Sonar Pro.
