Allows listing containers, viewing logs, and restarting or stopping Docker containers.
Provides Git operations such as status, branch, commit history, diff, push/pull, and stash management.
Supports cloning repositories, creating pull requests and issues, and listing branches.
Enables executing SQL queries, showing schema, describing tables, and listing databases.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@DevPilot MCP show me modified files"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
DevPilot is an AI agent built on top of the Model Context Protocol (MCP). It uses an LLM to reason about a user's request, invokes the appropriate MCP tools, observes their results, and continues this reasoning loop until it can produce a final answer.
User
│
▼
Conversation Manager
│
▼
Tool Registry (MCP)
│
list_tools() + inputSchema
│
▼
LLM (Planner)
│
┌───────────────┴────────────────┐
│ │
▼ ▼
Tool Call(s) Final Answer
│ │
▼ ▼
Executor Exit
│
▼
Tool Result(s)
│
▼
Conversation Manager
│
└───────────────► Back to LLMMaintains the complete conversation history.
Stores:
User messages
Assistant messages
Tool calls
Tool results
This conversation is sent to the LLM on every iteration.
Related MCP server: My Coding Buddy MCP Server
Discovers available MCP tools.
Provides:
Tool name
Description
Input schema
Required arguments
Example:
git_status
docker_logs
fs_read_file
docker_ps
...The LLM acts as the planner.
It receives:
Conversation history
Available tools
Tool schemas
It decides whether to:
Call one or more tools
Ask for clarification
Return the final answer
Executes tool calls returned by the LLM.
Input:
ToolCallOutput:
ToolResultUser
│
▼
Conversation Manager
│
▼
LLM
│
├── Final Answer ─────────────► Return Response
│
└── Tool Call
│
▼
Executor
│
▼
Tool Result
│
▼
Conversation Manager
│
▼
LLMThe cycle continues until the LLM returns a final answer.
User:
Why is my backend container failing?↓
LLM
Call docker_ps()↓
Executor
backend -> exited
db -> running↓
Conversation Updated
↓
LLM
Call docker_logs(container="backend")↓
Executor
panic:
database connection refused↓
Conversation Updated
↓
LLM
Final Answer
The backend container exits because it cannot connect to the database.The LLM can return one of the following responses.
{
"status": "tool_call",
"tool_calls": [
{
"tool_name": "docker_logs",
"tool_args": {
"container": "backend"
}
}
]
}{
"status": "final_answer",
"answer": "The backend container exits because it cannot connect to the database."
}{
"status": "clarification",
"answer": "Which repository do you want to inspect?"
}{
"status": "error",
"answer": "Unable to execute the requested tool."
}conversation.add_user(prompt)
tools = registry.list_tools()
while True:
response = llm.chat(
conversation,
tools
)
if response.status == "final_answer":
return response.answer
if response.status == "clarification":
return response.answer
tool_results = executor.execute(
response.tool_calls
)
conversation.add_tool_calls(
response.tool_calls
)
conversation.add_tool_results(
tool_results
)LLM is responsible for reasoning.
MCP tools are responsible for execution.
Executor never makes decisions.
Conversation history is preserved across iterations.
Tool execution is deterministic.
The agent loops until a final answer is produced.
This server cannot be installed
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
A comprehensive Model Context Protocol (MCP) server that enables AI assistants to interact with yo…
User-owned memory for AI agents, Copilot, Claude, IDEs, CLIs, and chat apps over remote MCP.
A Model Context Protocol server for Wix AI tools
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/walianitin/Dev-Mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server