embrace-raw-readonly
Embrace 只读全量数据 MCP
此本地 stdio MCP 保留官方 Embrace MCP 响应,包括 structuredContent,而不是仅返回网关摘要。
它不使用 Dashboard cookie,也不访问未记录的 Dashboard 端点。官方 Embrace MCP 目前公开聚合会话工具,以及详细的日志/崩溃/网络/span 结果;它不公开任意的 User Timeline 检索。
运行
父 MCP 进程必须继承服务账号令牌:
export EMBRACE_API_TOKEN='...'
uv run server.py不要将令牌作为工具参数传递,也不要将其提交到文件中。
Related MCP server: opencode-codex-mcp
工具
embrace_list_readonly_tools返回远程工具模式。embrace_call_readonly(tool_name, arguments)转发一个已列入白名单的工具,并返回完整的 JSON-RPC 响应。
服务器具有 22 个已发布 Embrace 工具的明确白名单,并拒绝任何其他内容。
技能
工作流技能位于 skills/embrace-readonly-full-data/SKILL.md。将该目录安装或复制到您的代理使用的技能目录中。该技能指示代理发现模式、保留结构化响应、串联详细工具,并避免未记录的 Dashboard 端点。
MCP 客户端配置
对于支持 stdio MCP 服务器的客户端:
{
"mcpServers": {
"embrace-raw-readonly": {
"command": "uv",
"args": ["run", "--project", "/path/to/embrace-readonly-mcp", "server.py"]
}
}
}客户端进程必须继承 EMBRACE_API_TOKEN。有关凭据处理,请参阅 docs/SECURITY.md。
Available Tools
2 toolsembrace_call_readonlyA
Call one published Embrace tool and return its complete JSON-RPC result.
This intentionally accepts arbitrary JSON arguments because the remote tool schemas evolve independently. Use embrace_list_readonly_tools first when the required arguments are unknown.
Inputs: tool_name (required), arguments. Returns: result
embrace_list_readonly_toolsA
Return the official Embrace read-only tool schemas as full JSON.
Inputs: no named parameters. Returns: result
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 2 tools
The two tools have clearly distinct purposes: one lists available tools, the other invokes a specific tool. There is no overlap or ambiguity in selecting between them.
Both tools start with 'embrace_' and include 'readonly', but one is 'list_readonly_tools' (verb_noun) while the other is 'call_readonly' (verb_adjective). The pattern is mostly consistent but not perfectly parallel.
With only two tools, the surface is thin. However, this is a deliberate thin proxy layer for a remote readonly API, so the minimalism is justified, but it still falls below the typical well-scoped range.
The two tools form a complete interface for interacting with the remote readonly toolset: listing available schemas and calling them. There are no missing operations within the stated scope of acting as a proxy.
Maintenance
Related MCP Connectors
Enable secure connectivity between Sentry issues and debugging data, and LLM clients, using a Model Context Protocol (MCP) server.
Analytics and debugging for your MCP server — explore usage and sessions, then root-cause errors.
MCP server for Statsig API - interact with Statsig's feature flags, experiments, and analytics
Read-only MCP access to sessions, funnels, campaigns, errors, live visitors, and anomalies.
Related MCP Servers
- AlicenseBqualityDmaintenanceEnables diagnosing, testing, benchmarking, and deliberately controlling LM Studio through a local-only MCP stdio server, with mutations disabled by default and evidence-based capability verification.18MIT
- FlicenseNot gradedqualityCmaintenanceEnables MCP-compatible hosts such as OpenCode to drive the Codex CLI through codex app-server over stdio, exposing tools to run prompts, inspect status, list threads, and interrupt running turns.-
- FlicenseNot gradedqualityBmaintenanceEnables MCP clients such as Claude Desktop or Claude Code to drive the local daemon through stdio, listing agents and sessions, running agents, and reporting costs.-
- AlicenseNot gradedqualityCmaintenanceEnables locally configured MCP servers to be called over stdio and mixed with xAI server-side tools, without relying on hosted xAI MCP.MIT