STONEWAY DOCUMENTATION
Give every AI agent the same you. Persistent, user-owned identity, projects, and context across all your coding tools.
1. Quick Start
StoneWay acts as a cloud-backed memory vault and bio generator for all your AI coding agents. One single bearer token (sw_...) coordinates memory across all tools on your computer.
Standing Rule for Agents
AI agents don't reliably discover or call tools by themselves. Add this single standing instruction to your agent's global or project rules files (CLAUDE.md, AGENTS.md, .cursor/rules, or GEMINI.md):
# StoneWay Agent Memory Instruction Before answering anything about me, my projects, or my preferences, call StoneWay's get_profile_context. After meaningful work, call append_note with what changed.
2. Connect Every Agent You Use
Choose your client below and paste the configuration into the corresponding settings file.
Claude Desktop
File: %APPDATA%\Claude\claude_desktop_config.json (Windows) or ~/Library/Application Support/Claude/claude_desktop_config.json (macOS).
{
"mcpServers": {
"stoneway": {
"command": "npx",
"args": ["-y", "stoneway-mcp"],
"env": {
"STONEWAY_TOKEN": "sw_your_token_here",
"STONEWAY_API_URL": "https://stonewaymd.vercel.app/api/v1"
}
}
}
}Claude Code CLI
Run this command directly in your terminal:
claude mcp add stoneway npx -y stoneway-mcp --env STONEWAY_TOKEN=sw_your_token_here
Cursor IDE
File: .cursor/mcp.json (per-repo) or Cursor Settings → Features → MCP:
{
"mcpServers": {
"stoneway": {
"command": "npx",
"args": ["-y", "stoneway-mcp"],
"env": {
"STONEWAY_TOKEN": "sw_your_token_here",
"STONEWAY_API_URL": "https://stonewaymd.vercel.app/api/v1"
}
}
}
}VS Code (GitHub Copilot)
File: .vscode/mcp.json:
{
"servers": {
"stoneway": {
"command": "npx",
"args": ["-y", "stoneway-mcp"],
"env": {
"STONEWAY_TOKEN": "sw_your_token_here"
}
}
}
}Windsurf & Cline
Add to your Cline / Windsurf MCP settings file:
{
"mcpServers": {
"stoneway": {
"command": "npx",
"args": ["-y", "stoneway-mcp"],
"env": {
"STONEWAY_TOKEN": "sw_your_token_here"
}
}
}
}Codex CLI
File: config.toml:
[mcp_servers.stoneway]
command = "npx"
args = ["-y", "stoneway-mcp"]
env = { STONEWAY_TOKEN = "sw_your_token_here" }Gemini CLI
File: settings.json:
{
"mcpServers": {
"stoneway": {
"command": "npx",
"args": ["-y", "stoneway-mcp"],
"env": {
"STONEWAY_TOKEN": "sw_your_token_here"
}
}
}
}Remote HTTP Server (Claude.ai & ChatGPT Cloud Agents)
Cloud agents cannot run terminal binaries on your computer. Point them to your hosted Streamable HTTP MCP URL:
https://stonewaymd.vercel.app/mcpBearer sw_your_token_herePython FastMCP / Native Stdio
If you prefer Python directly or want to test with the official MCP Inspector:
# Test with MCP Inspector (>= 0.14.1) npx @modelcontextprotocol/inspector py -m stoneway_mcp.server # Or direct Python command python -m stoneway_mcp.server
3. Daily Vibecoder Routine
- Instant Bios: Tell your agent
"Write my X bio"or"Draft my GitHub headline". The agent callsget_biousing strictly verified public facts. - Session Start: Your agent calls
get_profile_contextand automatically matches your preferred frameworks and conventions. - Session End: Tell your agent
"Log what we built". It callsappend_notewithout touching your code files. - Weekly Sync: Open /app/editor, skim your markdown notes, and click Sync Now on your GitHub connector.
- Resume Export: Click
JSON Resumein the editor or fetch/api/user/resumeto export your profile into standard JSON Resume format.
4. Available Tools Reference
get_profile_contextRetrieves structured StoneWay.json and StoneWay.md with reconciliation guidance. Call this first in any coding session.
update_profile_contextOptimistic-locking update to structured StoneWay.json with zero-data-loss overflow capture.
append_noteAppends timestamped notes, commit summaries, or architectural decisions to StoneWay.md.
get_bioGenerates platform-tailored bios (github, x, linkedin, devpost) filtered strictly to public fields.
trigger_external_syncTriggers on-demand synchronization for connected integrations (GitHub, npm, Notion).
export_json_resumeExports profile in standardized JSON Resume schema for interoperability.