从任何支持 MCP 的客户端运行 Sentrul 研究、合规性和追踪查询代理。 通过一个网关提供 257 个生产级代理。
版本: 0.1.0 | SDK: @modelcontextprotocol/sdk@1.29.0 | 许可证: MIT
将此代码片段添加到 ~/Library/Application Support/Claude/claude_desktop_config.json 中：
{
"mcpServers": {
"sentrul": {
"command": "npx",
"args": ["-y", "sentrul-mcp-server"],
"env": { "SENTRUL_API_KEY": "optional — free tier auto-provisions" }
}
}
}重启 Claude Desktop。服务器在首次使用时会自动配置免费层级密钥 — 无需注册。
Related MCP server: enterprise-knowledge-integrator
sentrul_research获取 URL 并对内容运行 Sentrul 的研究代理。 返回摘要、实体、声明以及公开的 Langfuse 追踪 URL。
示例提示词
"Research https://example.com/report and summarise the key claims"
"Do a full-depth research pass on https://arxiv.org/abs/2401.00001"
sentrul_compliance_scan扫描文本以查找常见的合规性模式（PII、PCI、HIPAA、GDPR 数据）。 返回调查结果及严重程度。非法律建议。
示例提示词
"Scan this document for PII and PCI violations: [paste text]"
"Check the following privacy policy for GDPR data patterns"
sentrul_trace_lookup通过 ID 检索公开的 Langfuse 追踪。 返回用于检查的结构化追踪负载。
示例提示词
"Fetch trace clm8abc123xyz from Sentrul"
"Show me the observation tree for trace ID abc-def-456"
变量 | 必需 | 默认值 | 描述 |
| 否 | 自动配置 | 您的 Sentrul API 密钥。留空以使用免费层级。 |
| 否 |
| 覆盖 Sentrul 网关基础 URL（用于暂存/本地开发）。 |
如果未设置 SENTRUL_API_KEY，服务器将：
检查 ~/.sentrul/mcp-key 中是否存在缓存密钥。
如果不存在缓存，则调用 POST /v1/mcp/keys/anonymous 以自动配置限速的免费密钥。
将密钥写入 ~/.sentrul/mcp-key（模式 0600）以供将来运行使用。
您可以随时通过在 MCP 客户端配置中设置 SENTRUL_API_KEY 来进行升级。
无 Shell 执行 — 服务器从不运行 shell 命令或生成子进程。
无 PII 日志记录 — 环境变量和 API 密钥绝不会写入日志或标准输出。
只读 — 所有三个工具在启动时均为只读；不会写入用户文件或外部状态。
文件系统写入 — 仅写入 ~/.sentrul/mcp-key（密钥缓存）。除此之外不会在目录外写入任何内容。
供应链 — 发布使用强制 2FA 的 npm 账户 + sigstore 认证（请参阅可用时的 SECURITY.md）。
pnpm install
pnpm run build # tsc → dist/
pnpm test # vitest run
# Smoke test — should print 3 tools in the tools/list response:
echo '{"jsonrpc":"2.0","id":1,"method":"tools/list"}' | node dist/index.jsMIT 许可证 — 版权所有 (c) 2026 Sentrul
Scan text for common compliance patterns (PII, PCI, HIPAA, GDPR data). Returns findings + severity. Not legal advice.
| Name | Required | Description | Default |
|---|---|---|---|
| text | Yes | ||
| frameworks | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden of behavioral disclosure. It mentions the tool 'Returns findings + severity,' which gives some insight into output behavior, but lacks details on error handling, rate limits, authentication needs, or performance characteristics. For a scanning tool with no annotation coverage, this is insufficient.
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 highly concise and front-loaded, consisting of two efficient sentences that convey core functionality and a disclaimer. Every sentence earns its place with no wasted words, making it easy to parse quickly.
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 tool's complexity (scanning for compliance patterns), lack of annotations, and no output schema, the description is incomplete. It doesn't explain the structure of returned findings, severity levels, error cases, or limitations. For a tool with two parameters and behavioral uncertainty, more context is needed to be fully helpful.
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 has 0% description coverage, so the description must compensate. It implies the 'text' parameter is the input to scan and lists compliance frameworks (PII, PCI, HIPAA, GDPR) that relate to the 'frameworks' parameter. However, it doesn't fully explain parameter semantics, such as allowed values for 'frameworks' or format expectations for 'text.' The baseline is 3 due to some added meaning but incomplete coverage.
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 purpose: 'Scan text for common compliance patterns (PII, PCI, HIPAA, GDPR data).' It specifies the verb ('scan'), resource ('text'), and scope ('compliance patterns'). However, it doesn't explicitly differentiate from sibling tools like 'sentrul_research' or 'sentrul_trace_lookup,' which prevents a perfect score.
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 provides minimal usage guidance. It mentions what the tool does but offers no explicit advice on when to use it versus alternatives, nor does it mention prerequisites or exclusions. The disclaimer 'Not legal advice' is helpful but doesn't constitute usage guidelines for tool selection.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Fetch a URL and run Sentrul's research-agent on the content. Returns summary, entities, claims, and a public Langfuse trace URL.
| Name | Required | Description | Default |
|---|---|---|---|
| url | Yes | ||
| depth | No | quick |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden of behavioral disclosure. It mentions fetching a URL and running a research-agent, which implies external network calls and processing, but lacks details on permissions, rate limits, error handling, or what 'quick' vs. 'full' depth entails. It states the return values but not their format or potential side effects.
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, well-structured sentence that efficiently conveys the core action, inputs, and outputs without redundancy. It is front-loaded with the main purpose and appropriately sized for the tool's complexity.
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 tool has no annotations, no output schema, and 0% schema description coverage, the description is incomplete. It covers the basic purpose and outputs but lacks details on behavioral traits, parameter meanings, and usage context. For a tool with external dependencies and processing, more information is needed to be fully helpful.
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 description does not explicitly mention parameters, but with 0% schema description coverage, it must compensate. It implies the 'url' parameter by stating 'Fetch a URL' and hints at 'depth' through 'run Sentrul's research-agent on the content' (suggesting analysis depth). However, it doesn't detail the enum values ('quick', 'full') or their effects, leaving some semantic gaps.
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 purpose with specific verbs ('fetch', 'run') and resources ('URL', 'Sentrul's research-agent'), and distinguishes it from siblings by specifying the research-agent focus (vs. compliance_scan or trace_lookup). It explicitly lists the outputs (summary, entities, claims, Langfuse trace URL).
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 provides no guidance on when to use this tool versus the sibling tools (sentrul_compliance_scan, sentrul_trace_lookup). It mentions the tool's function but offers no context about alternatives, prerequisites, or exclusions, leaving the agent to infer usage scenarios.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Retrieve a public Langfuse trace by ID. Returns the structured trace payload for inspection in the client.
| Name | Required | Description | Default |
|---|---|---|---|
| trace_id | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden of behavioral disclosure. It states the tool retrieves a 'public' trace and returns a 'structured trace payload', but lacks details on permissions, rate limits, error handling, or what 'public' entails (e.g., access controls). This leaves significant gaps for a tool that likely involves data access.
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 appropriately sized with two sentences that are front-loaded and efficient. It avoids unnecessary details, though it could be slightly more structured (e.g., separating purpose from output).
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 tool's low complexity (1 parameter, no output schema, no annotations), the description is somewhat complete but lacks depth. It covers the basic action and output intent, but misses behavioral aspects like authentication needs or error cases, making it adequate but with clear gaps.
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 description adds minimal meaning beyond the input schema, which has 0% coverage. It implies 'trace_id' is used to look up a trace, but doesn't specify format, source, or constraints (e.g., length, pattern). With low schema coverage, the description doesn't fully compensate, so it meets the baseline for adding some context.
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 ('Retrieve') and resource ('public Langfuse trace by ID'), making the purpose specific and understandable. However, it doesn't explicitly differentiate from sibling tools like 'sentrul_compliance_scan' or 'sentrul_research', which might have overlapping domains but different functions.
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 provides no guidance on when to use this tool versus alternatives, such as the sibling tools. It mentions 'for inspection in the client' but doesn't clarify scenarios, prerequisites, or exclusions, leaving usage context implied at best.
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.
v0.1.0sentrul_compliance_scansentrul_researchsentrul_trace_lookupScored across 3 tools
Each tool has a clearly distinct purpose with no overlap. sentrul_compliance_scan analyzes text for compliance patterns, sentrul_research processes web content for research insights, and sentrul_trace_lookup retrieves specific trace data. An agent can easily differentiate between scanning, researching, and trace lookup functions.
All tool names follow a consistent 'sentrul_' prefix with descriptive suffixes (compliance_scan, research, trace_lookup). The naming pattern is uniform and predictable, using snake_case throughout without any deviations or mixed conventions.
With only 3 tools, the set feels thin for a server named 'sentrul-mcp-server', which suggests a broader scope for compliance and research operations. While each tool is distinct, the count is borderline low, potentially limiting functionality for agents working in this domain.
The tools cover scanning, research, and trace lookup, but there are notable gaps for a compliance/research domain. Missing operations include updating or managing scans, configuring research parameters, or handling trace modifications, which could lead to dead ends in agent workflows.
MCP tools for agents: web research, content extraction, email, DNS, and blockchain intelligence.
Compliance frameworks (SOC 2, ISO 27001, CMMC, NIST, more) delivered to AI agents as MCP tools.
A paid remote MCP for agent memory MCP, built to return verdicts, receipts, usage logs, and audit-re
100+ MCP tools for AI agents: content metadata, trade intelligence, business-expertise analysis.