Synesthetic MCP
Enables validation, diff generation, and backend population for Synesthetic assets using schema-compliant resources and tools. Serves as an MCP adapter that enforces schema compliance and integrates with the Synesthetic asset generation pipeline.
README
version: 0.1.0 owner: delk73 lastReviewed: 2025-09-12
Synesthetic MCP
Lightweight Model Context Protocol (MCP) server for Synesthetic assets.
This repo adapts the canonical synesthetic-schemas into MCP resources and tools.
- Resources: schemas, examples (read-only, from SSOT).
- Tools: validation, diff (RFC6902 patch ops), backend population.
- Guards: enforce schema compliance before persistence.
System Context
flowchart LR
FE["sdfk_fe_ts\n(TypeScript Frontend)"]
BE["sdfk-backend\n(Python API & CRUD)"]
MCP["synesthetic-mcp\n(Python MCP Adapter)"]
RLHF["sdfk_rlhf\n(RLHF Asset Generation)"]
SSOT["synesthetic-schemas\n(SSOT: Schemas + Codegen)"]
FE --> |Consumes API| BE
RLHF --> |Generates Assets via MCP| MCP
MCP --> |Validates + Persists| BE
SSOT --> |Submodule + Python pkg| MCP
SSOT --> |TypeScript types| FE
SSOT --> |Python models| BE
style MCP fill:#444444,stroke:#ffffff,stroke-width:2px
Quick Setup
Option A: Nix + Poetry (full dev shell)
# Enter reproducible Nix shell (Python + Poetry pre-installed)
nix develop
# Install deps into local .venv/
poetry install
# Run tests
pytest
Option B: Fast path (pinned requirements, no resolver rodeos)
# Clean and bootstrap fresh venv
./install.sh
# Activate venv
source .venv/bin/activate
# Run tests
pytest -q
# Lint / type-check
ruff check .
mypy mcp/
The
install.shscript installs fromrequirements.txtwith exact pinned versions for deterministic, fast environment setup.
Structure
synesthetic-mcp/
README.md
requirements.txt # Pinned dependencies (single source of truth for pip)
install.sh # Simple bootstrap script for reproducible env
docs/
mcp_spec.md # Full spec for MCP v1 (resources, tools, error model)
mcp/
__init__.py
core.py # schema/example discovery
validate.py # validation helpers
diff.py # JSON Patch generation
backend.py # proxy client to backend
stdio_main.py # MCP stdio adapter
http_main.py # optional FastAPI adapter
tests/
test_validate.py
test_diff.py
test_tools.py
fixtures/
valid_asset.json
invalid_asset.json
meta/
prompts/ # Canonical Codex/ChatGPT prompts used to spin up repo
Development
- MCP server is written in Python (>=3.11, tested on 3.11–3.13).
- Depends on generated Python models from
synesthetic-schemas. - Use FastAPI for HTTP adapter, or run via stdio for direct agent integration.
- Tests use pytest and golden fixtures.
- Reproducible dev environment provided via Nix flake; pip +
requirements.txtfor fast installs.
Spec
See docs/mcp_spec.md for the pinned v1 specification.
Status
✅ Spec pinned in docs/mcp_spec.md
🚧 Implementation scaffolding in progress
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。