Agent Loop MCP Server

All components are designed for modularity, minimalism, and functional programming style.

Local serverstdioPython

What is the Agent Loop MCP server?

Connect Agent Loop to Claude, Cursor or any other MCP client and it stops being a tab you switch to. All components are designed for modularity, minimalism, and functional programming style. The agent loop mcp server is what makes that connection.

What the server does

  • Conversational AI agent powered by Anthropic Claude or OpenAI GPT
  • Configurable AI provider and temperature — via environment variables
  • Pragmatic loop control — with iteration limits and completion detection
  • Tool execution with optional human confirmation (--safe mode)
  • Debug mode for transparency (--debug)
  • Custom tools support — with automatic discovery and display

Available tools

The toolset is worth reading before you wire it up, because it tells you what the integration is really for:

  • Drop — in Python modules** (pure functions, functional programming style)
  • Auto-discovered — Just place your .py file in agent_loop/tools/ (built-in) or ~/.config/agent-loop/tools/ (user tools)
  • bash — Execute bash commands
  • sympy — Perform symbolic mathematics operations using SymPy
  • cli_plot — Render advanced terminal charts and plots using plotext
  • filesystem — Read, create, update, append, delete files with UTF-8 encoding
  • list_dir — List the contents of a directory for quick file discovery
  • codebase_search — Semantic code search for relevant code snippets in the project
  • file_search — Fast fuzzy file search by filename or path fragment
  • grep_search — Search for exact strings or regex patterns in files
  • curl — Make HTTP requests using curl
  • project_inspector — Inspect the current project directory and preview source files

Credentials and setup notes

Configuration is passed through the environment: BRAVE_API_KEY, ANTHROPIC_API_KEY, OPENAI_API_KEY, JIRA_BASE_URL, JIRA_API_TOKEN, CONFLUENCE_BASE_URL, CONFLUENCE_API_TOKEN. Treat anything key-shaped as a real credential — scope it to the minimum the server needs, and rotate it if it ever lands in a shared config.

  • Python: >= 3.12 - Core Python dependencies: - anthropic >= 0.51.0 - halo >= 0.0.31 - mcp[cli] >= 1.9.2 - openai >= 1.79.0 - plotext >= 5.3.2 - python-dotenv >= 1.1.0 - requests >= 2.32.3 - sympy >= 1.14.0 - Recommended for installation: - uv (for fast dependency installation) - Optional/for full tool support: - Node.js (for some MCP server integrations

Installation

Setup follows the usual MCP pattern — install or clone the server, register it in your client's configuration file, restart the client. The configuration blocks on this page cover the common clients.

Where it fits

Plenty of developer tooling servers cover similar ground. The differences that matter in practice are scope of access and how much setup stands between you and a working tool call. Agent Loop's toolset — Drop, Auto-discovered, bash and 11 more — is a fair guide to whether it matches your workflow. It is maintained by AlessandroAnnini; worth a glance at recent repository activity before you build anything load-bearing on it.

We check each listing at SyncDev against the project's documentation before it goes live — if something here drifts out of date, it is a bug worth reporting.

Worth knowing first

  • It runs with your machine's permissions. That is convenient and also the reason to think about what you point it at before you approve a tool call.
  • With 14 tools registered it takes up a noticeable share of the context window; turn it off in projects that never touch Agent Loop.
  • Missing credentials fail quietly in some clients — if no tools show up, check the environment block first.
  • Keep per-call confirmation enabled while you learn its behaviour; it is the cheapest safeguard you have.

Available tools

ToolWhat it does
Dropin Python modules** (pure functions, functional programming style)
Auto-discoveredJust place your .py file in agent_loop/tools/ (built-in) or ~/.config/agent-loop/tools/ (user tools)
bashExecute bash commands
sympyPerform symbolic mathematics operations using SymPy
cli_plotRender advanced terminal charts and plots using plotext
filesystemRead, create, update, append, delete files with UTF-8 encoding
list_dirList the contents of a directory for quick file discovery
codebase_searchSemantic code search for relevant code snippets in the project
file_searchFast fuzzy file search by filename or path fragment
grep_searchSearch for exact strings or regex patterns in files
curlMake HTTP requests using curl
project_inspectorInspect the current project directory and preview source files
kubectlRun kubectl commands to interact with a Kubernetes cluster
aws_cliRun AWS CLI v2 read-only commands to interact with AWS services

How to install the Agent Loop MCP server

{
  "mcpServers": {
    "brave-search": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-brave-search"],
      "env": { "BRAVE_API_KEY": "..." }
    },
    "mcp-obsidian": {
      "command": "npx",
      "args": ["-y", "mcp-obsidian", "/path/to/obsidian-vault/"]
    }
  }
}

Configuration as documented by the project. Restart the client after saving.

Configuration

  • Python: >= 3.12 - Core Python dependencies: - anthropic >= 0.51.0 - halo >= 0.0.31 - mcp[cli] >= 1.9.2 - openai >= 1.79.0 - plotext >= 5.3.2 - python-dotenv >= 1.1.0 - requests >= 2.32.3 - sympy >= 1.14.0 - Recommended for installation: - uv (for fast dependency installation) - Optional/for full tool support: - Node.js (for some MCP server integrations
VariableDescriptionRequired
BRAVE_API_KEYCredential the server authenticates with.Yes
ANTHROPIC_API_KEYCredential the server authenticates with.Yes
OPENAI_API_KEYCredential the server authenticates with.Yes
JIRA_BASE_URLEndpoint or connection string the server talks to.Yes
JIRA_API_TOKENCredential the server authenticates with.Yes
CONFLUENCE_BASE_URLEndpoint or connection string the server talks to.Yes
CONFLUENCE_API_TOKENCredential the server authenticates with.Yes

Example prompts to try

  • Use Agent Loop to Drop.
  • Use Agent Loop to Auto-discovered.
  • Use Agent Loop to bash.

Frequently asked questions

It connects Agent Loop to MCP-compatible AI assistants such as Claude and Cursor, exposing 14 tools (Drop, Auto-discovered, bash, and more) that the assistant can call on your behalf. Instead of copying data back and forth by hand, the assistant works with Agent Loop directly.