AI-Powered Context Protocol Server (AIO-MCP)

This robust Model Context Protocol (MCP) server implementation integrates deeply with major enterprise and public tools—GitLab, Jira, Confluence, YouTube, and Google Maps—to enable intelligent, context-aware search and execution functionalities within software development lifecycles.

System Requirements

Go Language runtime, version 1.23.2 or newer.
Valid access credentials (tokens, keys) for all intended external integrations.

Deployment Strategies

Automated Provisioning via Smithery

For streamlined, interactive setup via the Claude Desktop environment, utilize the dedicated Smithery installer:

npx -y @smithery/cli install @athapong/aio-mcp --client claude

Note: Smithery manages configuration prompting and environment variable scaffolding automatically.

Manual Compilation and Installation (Go Toolchain)

To install directly using go install, ensure your Go environment variables direct binaries to a PATH-accessible location. This typically requires configuring GOPATH and ensuring $GOPATH/bin is included in your system's PATH.

Configuration Guidance (Shell Setup):

Unix-like Systems (Linux/macOS):

Add the following lines to your shell startup file (e.g., ~/.bashrc, ~/.zshrc):

export GOPATH=$HOME/go
export PATH="$PATH:$GOPATH/bin"

Apply changes with source ~/.bashrc (or equivalent).

Windows (PowerShell):

Use PowerShell cmdlets to set user environment variables:

[Environment]::SetEnvironmentVariable("GOPATH", "$env:USERPROFILE\go", "User")
[Environment]::SetEnvironmentVariable("PATH", "$env:PATH;$env:USERPROFILE\go\bin", "User")

Windows (Command Prompt):

setx GOPATH "%USERPROFILE%\go"
setx PATH "%PATH%;%USERPROFILE%\go\bin"

Verification: Confirm settings with go env GOPATH and checking the relevant $PATH output.

Installation Command:

go install github.com/athapong/aio-mcp@latest

Configuration File Setup

Environment Configuration (.env file): Populate this file with necessary secrets and configuration parameters. If left empty, all tools default to being disabled or null unless explicitly configured otherwise: env ENABLE_TOOLS= QDRANT_HOST= ATLASSIAN_HOST= ATLASSIAN_EMAIL= GITLAB_HOST= GITLAB_TOKEN= BRAVE_API_KEY= ATLASSIAN_TOKEN= GOOGLE_AI_API_KEY= PROXY_URL= OPENAI_API_KEY= OPENAI_EMBEDDING_MODEL= DEEPSEEK_API_KEY= QDRANT_PORT= GOOGLE_TOKEN_FILE= GOOGLE_CREDENTIALS_FILE= QDRANT_API_KEY= USE_OLLAMA_DEEPSEEK= ENABLE_SSE= SSE_ADDR= SSE_BASE_PATH=
Client Integration (Claude Desktop JSON Example): Configure the client to launch the server, specifying environment file paths and enabling Server-Sent Events (SSE):

json{claude_desktop_config.json} { "mcpServers": { "aio-mcp": { "command": "aio-mcp", "args": ["-env", "/path/to/.env", "-sse", "-sse-addr", ":8080", "-sse-base-path", "/mcp"], } } }

Or, configure environment overrides directly in the client JSON:

```json { "mcpServers": { "aio-mcp": { "command": "aio-mcp", "env": { "ENABLE_TOOLS": "", "OPENAI_BASE_URL": "", "GOOGLE_AI_API_KEY": "", "GITLAB_TOKEN": "", "GITLAB_HOST": "", "QDRANT_HOST": "", "QDRANT_API_KEY": "",
```
    "PROXY_URL": "",
    "OPENAI_API_KEY": "",
    "GOOGLE_TOKEN_FILE": "",
    "GOOGLE_CREDENTIALS_FILE": "",

    "ATLASSIAN_TOKEN": "",
    "BRAVE_API_KEY": "",
    "QDRANT_PORT": "",
    "ATLASSIAN_HOST": "",
    "ATLASSIAN_EMAIL": "",

    "USE_OPENROUTER": "", // "true" if utilizing openrouter for reasoning, default false
    "DEEPSEEK_API_KEY": "", // Deepseek credentials for reasoning
    "OPENROUTER_API_KEY": "", // OpenRouter credentials for reasoning
    "DEEPSEEK_API_BASE": "", // Custom Deepseek base endpoint
    "USE_OLLAMA_DEEPSEEK": "", // "true" for local Ollama Deepseek execution, default false
    "OLLAMA_URL": "" // Default: http://localhost:11434
  }
}
```
} } ```

Operational Modes

The AIO-MCP Server supports two distinct communication paradigms:

Standard I/O (Default): Primary mode for interaction with the Claude Desktop client, relying on stdin/stdout streams.
Server-Sent Events (SSE) Mode: Operates an independent HTTP listener for continuous, real-time data transmission, suitable for network-accessible clients or web applications.

Activating SSE Mode

SSE functionality can be toggled via:

Command Line Arguments: bash aio-mcp -sse -sse-addr ":8080" -sse-base-path "/mcp"
Environment Variables (in .env file): ENABLE_SSE=true SSE_ADDR=:8080 SSE_BASE_PATH=/mcp

When active, the server exposes these HTTP endpoints: * Event Stream: {SSE_BASE_PATH}/sse (Default: /mcp/sse) * Message Submission: {SSE_BASE_PATH}/message (Default: /mcp/message)

Tool Activation Control

The ENABLE_TOOLS environment variable governs which functional toolsets are available. If this variable is unset or empty, all integrated tools are initialized. Provide a comma-separated list of tool groups to restrict functionality.

Available Tool Groups:

gemini: Access to Google Gemini search capabilities.
fetch: Generic HTTP content retrieval.
brave_search: Web querying via Brave Search.
google_maps: Geospatial querying and data retrieval.
confluence: Atlassian documentation interactions.
youtube: Video information and transcript retrieval.
jira: Atlassian project and issue management.
gitlab: Source code repository management and CI/CD inspection.
script: Sandboxed command execution environment.
rag: Retrieval-Augmented Generation (Vector Memory) indexing and querying.
deepseek: Advanced reasoning utilizing Deepseek models (requires USE_OLLAMA_DEEPSEEK=true for local execution on default model deepseek-r1:8b).

Integrated Tool Definitions

Google Calendar Utilities

calendar_create_event: Schedules a new calendar appointment.
calendar_list_events: Retrieves a schedule of upcoming meetings.
calendar_update_event: Modifies details of an existing engagement.
calendar_respond_to_event: Accepts, declines, or tentatively accepts an invitation.

Confluence Documentation Services

confluence_search: Executes a query against the Atlassian knowledge base using CQL.
confluence_get_page: Retrieves the raw storage format content for a specified document ID.
confluence_create_page: Authors a new page within a specified Confluence space.
confluence_update_page: Overwrites content or metadata of an existing page.
confluence_compare_versions: Diffs two historical states of a page.

Deepseek Reasoning Engine

deepseek_reasoning: Employs Deepseek models for complex, multi-faceted analytical tasks, strategic planning, and detailed problem decomposition, benefiting from structured context input.

Data Fetching

get_web_content: Fetches and returns raw text content from any accessible URI (HTTP/HTTPS).

Google Chat Integration

gchat_list_spaces: Enumerates accessible Chat rooms/groups.
gchat_send_message: Posts a text notification to a specified room or direct conversation.

AI Web Search (Gemini Integration)

ai_web_search: Leverages Google AI Search APIs for up-to-date information retrieval, requiring a descriptive query and usage context.

GitLab Repository Management

gitlab_list_projects: Lists repositories within a specified group ID.
gitlab_get_project: Fetches metadata for a repository via its path.
gitlab_list_mrs: Queries open, closed, or merged Merge Requests for a project.
gitlab_get_mr_details: Retrieves granular data for a specific MR.
gitlab_create_MR_note: Adds commentary to an existing Merge Request.
gitlab_get_file_content: Downloads the raw text contents of a file at a specific revision.
gitlab_list_pipelines: Shows status history for project CI/CD pipelines.
gitlab_list_commits: Retrieves commit history between specified dates/references.
gitlab_get_commit_details: Examines specifics of a single commit SHA.
gitlab_list_user_events: Tracks activity log for a specified GitLab user.
gitlab_list_group_users: Lists all members associated with a GitLab group.
gitlab_create_mr: Initiates a new Merge Request between branches.
gitlab_clone_repo: Executes a local clone or sync operation for a specified repository path.

Gmail Client Operations

gmail_search: Executes standard Gmail search syntax to locate messages.
gmail_move_to_spam: Marks specified messages as junk.
gmail_create_filter: Establishes automated rules for incoming mail.
gmail_list_filters: Displays all currently active mail processing rules.
gmail_list_labels: Shows all user-defined labels.
gmail_delete_filter: Removes an established filtering rule.
gmail_delete_label: Removes a specified label definition.

Jira Project Tracking

jira_get_issue: Fetches comprehensive details, status, and transition options for an issue key.
jira_search_issue: Finds issues matching a specified JQL expression.
jira_list_sprints: Reports on ongoing and scheduled sprints for a board.
jira_create_issue: Provisions a new work item (Task, Bug, Story, etc.).
jira_update_issue: Applies targeted modifications to an existing ticket.
jira_list_statuses: Lists permissible workflow states for a project.
jira_transition_issue: Moves an issue to a subsequent workflow state using a known transition ID.

RAG Memory Management (Vector Indexing)

RAG_memory_index_content: Embeds provided plain text content into a named vector collection.
RAG_memory_index_file: Indexes the content of a local file into memory storage.
RAG_memory_create_collection: Establishes a new, empty collection for vector storage.
RAG_memory_delete_collection: Removes an entire vector collection.
RAG_memory_list_collections: Lists all currently persisted vector collections.
RAG_memory_search: Performs a similarity search against indexed data within a collection.
RAG_memory_delete_index_by_filepath: Removes entries associated with a specific source file.

System Execution

execute_comand_line_script: Executes shell commands in a restricted, sandboxed environment with timeout controls. Supports various interpreters (e.g., /bin/sh, Python).

Brave Web Search

web_search: Queries the Brave Search index, allowing specification of result count and region.

Reflective Processing & Planning

sequentialthinking: A highly detailed meta-tool supporting iterative, reflective problem-solving. It encourages breaking down complex tasks, generating hypotheses, self-critique, revision, and adaptation across multiple thought steps (thoughtNumber). Use it for deep planning where linear execution is insufficient.
sequentialthinking_history: Retrieves the recorded analytical trace of a previous reflective session.
tool_manager: Administers the enablement/disablement status of other integrated tools.
tool_use_plan: Generates an optimal sequence of tool calls to satisfy a high-level user request.

YouTube Content Access

youtube_transcript: Extracts the textual transcript data from a specified video ID.
youtube_update_video: Modifies metadata (title, description, keywords, category) for a user-owned video.
youtube_get_video_details: Fetches comprehensive metadata for a given video identifier.
youtube_list_videos: Enumerates videos associated with a user's channel.

WIKIPEDIA CONTEXT: Cloud computing, defined by ISO, is a model providing on-demand access to a scalable, shared pool of configurable computing resources. NIST standardized five core characteristics: On-demand self-service, Broad network access, Resource pooling, Rapid elasticity, and Measured service. The foundational concepts trace back to 1960s time-sharing systems; the modern 'cloud' metaphor materialized in the mid-1990s.

Cognito-Interface-Framework

Author

athapong

Quick Info

Actions

Tags