personal-mcp
Provides web search and news search capabilities using DuckDuckGo, enabling retrieval of information from the web.
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
personal-mcp
Personal concierge MCP server — productivity tools and FastMCP transforms for SLM orchestrator RL training.
What this is
A FastMCP server exposing 22 tools that an LLM/SLM orchestrator can call: calendar, to-do, notes, web search, calculator, scratchpad, task queue, and sandboxed code execution. The server is designed to be:
Safe by default — all state is in-memory and synthetic; no real user data is touched.
RL-friendly — tools return structured, deterministic outputs; destructive actions require explicit confirmation; errors are returned as
{"error": "..."}dicts (not exceptions) so the agent learns to handle malformed input.Transform-ready — three FastMCP transforms are wired in by default (ToolTransform for tags, BM25SearchTransform for on-demand discovery, optional CodeMode for the "write Python to orchestrate tools" pattern).
Related MCP server: production-grade-mcp-agentic-system
Install
pip install -e ".[dev]"This installs fastmcp[code-mode] and ddgs, plus dev tools (pytest, ruff, bandit).
Run
# stdio (for Claude Desktop / MCP Inspector)
fastmcp run server.py
# Streamable HTTP (for ART, DeepAgents, remote clients)
fastmcp run server.py --transport streamable-http --port 8000
# or just:
python server.pyTwo runtime modes
Toggle with the CODE_MODE_ENABLED env var:
default (
CODE_MODE_ENABLED=0): clients see 3 tools —search_tools,call_tool,reset_state— and discover the rest on demand via BM25.code mode (
CODE_MODE_ENABLED=1): clients see 4 meta-tools —tags,search,get_schema,execute— and write Python that chainscall_tool()invocations inside a sandboxed Monty interpreter.
Tools
Category | Tool | Purpose |
State |
| Clear all server state (call between training episodes) |
Calendar |
| Calendar management |
To-do |
| Simple to-do tasks |
Notes |
| Short notes with tag search |
Search |
| DuckDuckGo web + news search |
Math |
| Safe AST-sandboxed arithmetic |
Memory |
| Working memory for multi-step plans |
Planning |
| Explicit task queue for plan decomposition |
Code |
| Sandboxed Python subprocess (5s default timeout) |
Test
pytest test_app.py -v
ruff check server.py test_app.py
bandit -r server.pyTool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Related MCP Connectors
MCP server for your apps' tools and custom tools, plus hosted AI agents and approval-gated workflows
214 tools for AI agents over one MCP endpoint: image, video, audio, 3D, data, storage, mail, GPUs.
Stateless agentic tools over MCP: concept extraction, long-context, knowledge graph, planning.
Email inboxes, webhook URLs, mock APIs and 170+ tools for coding agents over one MCP connection.
Related MCP Servers
- AlicenseNot gradedqualityNot gradedmaintenanceA lightweight and fast MCP server that enables AI agents to efficiently discover and execute tools through progressive disclosure, minimizing context consumption while supporting safe code execution in external environments.14 npm-
- AlicenseNot gradedqualityDmaintenanceA production-grade MCP server designed for multi-tenant, authenticated, and observable AI agent systems, enabling secure tool execution across heterogeneous data sources.68MIT
- AlicenseCqualityDmaintenanceA Model Context Protocol server exposing 58 online tools (crypto, data, image, text, etc.) and workflow execution, enabling MCP clients like Claude Desktop to invoke them via natural language.588 npmMIT
- FlicenseNot gradedqualityDmaintenanceA multi-tool task agent MCP server with file search, SQLite query, calculator, and report writing tools. Enables Claude Code, Claude Desktop, or Cursor to control the same tools used by the agent, with guardrails for safety.-