### Start Unified Server (Bash)

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Examples of starting the MCP unified server using the 'uvx' command. Includes basic usage, restricting tool search, SSE transport, Docker deployment, and debugging with MCP Inspector.

```bash
export ACI_API_KEY="your_aci_api_key_here"

# Basic usage (stdio transport)
uvx aci-mcp unified-server \
  --linked-account-owner-id "user_123"
```

```bash
# Restrict tool search to apps explicitly allowed for this API key
uvx aci-mcp unified-server \
  --linked-account-owner-id "user_123" \
  --allowed-apps-only
```

```bash
# SSE transport for remote/HTTP clients
uvx aci-mcp unified-server \
  --linked-account-owner-id "user_123" \
  --transport sse \
  --port 8080
```

```bash
# Docker
docker run --rm -i \
  -e ACI_API_KEY=your_aci_api_key_here \
  aci-mcp unified-server \
  --linked-account-owner-id "user_123"
```

```bash
# Debug with MCP Inspector
npx @modelcontextprotocol/inspector \
  uvx aci-mcp unified-server \
  --linked-account-owner-id "user_123"
```

--------------------------------

### Start ACI MCP Apps Server (stdio)

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Starts the apps-server with stdio transport, exposing tools from specified ACI.dev apps. Requires ACI_API_KEY and linked-account-owner-id.

```bash
# Set your API key
export ACI_API_KEY="your_aci_api_key_here"

# Run via uvx (stdio transport — default, for Claude Desktop / Cursor)
uvx aci-mcp apps-server \
  --apps "BRAVE_SEARCH,GMAIL" \
  --linked-account-owner-id "user_123"
```

--------------------------------

### Start VibeOps Specialized MCP Server

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Starts a specialized MCP server for VibeOps workflows, targeting AI-assisted project scaffolding. It uses a separate API key and connects to a specific controller URL.

```bash
export VIBEOPS_API_KEY="your_vibeops_api_key_here"

# Start the vibeops server (stdio only, no transport options)
uvx aci-mcp vibeops-server

# The server exposes:
# - GET_PROJECT_STATE: retrieves current GitLab, Vercel, and Supabase project state
#   and returns rich guidance on Next.js/TypeScript best practices.
# - All dynamically fetched VibeOps functions from /v1/functions/search
```

--------------------------------

### Start ACI MCP Apps Server (SSE)

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Starts the apps-server with SSE transport on a specified port, exposing tools from specified ACI.dev apps. Requires ACI_API_KEY and linked-account-owner-id.

```bash
# Run with SSE transport (for network/HTTP clients) on port 9000
uvx aci-mcp apps-server \
  --apps "BRAVE_SEARCH,GMAIL,GITHUB" \
  --linked-account-owner-id "user_123" \
  --transport sse \
  --port 9000
```

--------------------------------

### Install uv Package Manager

Source: https://github.com/aipotheosis-labs/aci-mcp/blob/main/README.md

Installs the uv package manager, which is used to run MCP servers locally. Ensure you have pip installed.

```bash
# Install uv if you don't have it already
curl -sSf https://install.pypa.io/get-pip.py | python3 - 
pip install uv
```

--------------------------------

### unified-server Command

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Starts an MCP server that exposes meta-tools for agents to discover and invoke functions on ACI.dev.

```APIDOC
## `unified-server` Command

Starts an MCP server that exposes two meta-tools (`ACI_SEARCH_FUNCTIONS`, `ACI_EXECUTE_FUNCTION`) plus a documentation search tool (`ACI_SEARCH_DOCS`). This server is designed for agents that need to autonomously discover and invoke any function available on ACI.dev without being pre-configured with specific apps.

```bash
export ACI_API_KEY="your_aci_api_key_here"

# Basic usage (stdio transport)
uvx aci-mcp unified-server \
  --linked-account-owner-id "user_123"

# Restrict tool search to apps explicitly allowed for this API key
uvx aci-mcp unified-server \
  --linked-account-owner-id "user_123" \
  --allowed-apps-only

# SSE transport for remote/HTTP clients
uvx aci-mcp unified-server \
  --linked-account-owner-id "user_123" \
  --transport sse \
  --port 8080

# Docker
docker run --rm -i \
  -e ACI_API_KEY=your_aci_api_key_here \
  aci-mcp unified-server \
  --linked-account-owner-id "user_123"

# Debug with MCP Inspector
npx @modelcontextprotocol/inspector \
  uvx aci-mcp unified-server \
  --linked-account-owner-id "user_123"
```
```

--------------------------------

### ACI_EXECUTE_FUNCTION Tool Usage (Python)

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

This Python code illustrates the schema and example usage for the ACI_EXECUTE_FUNCTION tool. It enables agents to execute ACI.dev functions by name and arguments, supporting multi-user overrides.

```python
# The tool is exposed to MCP clients as:
# {
#   "name": "ACI_EXECUTE_FUNCTION",
#   "description": "Execute a function on ACI.dev by name.",
#   "inputSchema": {
#     "type": "object",
#     "properties": {
#       "function_name": { "type": "string" },
#       "function_arguments": { "type": "object" }
#     },
#     "required": ["function_name", "function_arguments"]
#   }
# }

# Example agent call (tools/call):
# name: "ACI_EXECUTE_FUNCTION"
# arguments: {
#   "function_name": "GMAIL__SEND_EMAIL",
#   "function_arguments": {
#     "to": "alice@example.com",
#     "subject": "Hello from ACI",
#     "body": "This email was sent by an AI agent via ACI MCP."
#   },
#   "aci_override_linked_account_owner_id": "other_user_456"  # optional multi-user override
# }

```

--------------------------------

### Run SSE Server with Starlette and Uvicorn

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Sets up an SSE server using SseServerTransport and integrates it with a Starlette application. Requires uvicorn to serve.

```python
async def run_sse_async(server: Server, host: str, port: int):
    sse = SseServerTransport("/messages/")

    async def handle_sse(request):
        async with sse.connect_sse(request.scope, request.receive, request._send) as streams:
            await server.run(streams[0], streams[1], server.create_initialization_options())

    starlette_app = Starlette(
        routes=[
            Route("/sse", endpoint=handle_sse),
            Mount("/messages/", app=sse.handle_post_message),
        ],
    )
    config = uvicorn.Config(starlette_app, host=host, port=port, log_level="debug")
    await uvicorn.Server(config).serve()
```

--------------------------------

### Show ACI MCP CLI Help

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Displays all available commands for the aci-mcp CLI. Ensure the ACI_API_KEY environment variable is set before running.

```bash
uvx aci-mcp --help
```

--------------------------------

### Build and Run MCP Servers with Docker

Source: https://github.com/aipotheosis-labs/aci-mcp/blob/main/README.md

Builds the Docker image for aci-mcp and provides commands to run both the unified and apps MCP servers. Ensure you replace placeholders like <ACI_API_KEY> and <LINKED_ACCOUNT_OWNER_ID>.

```bash
# Build the image
docker build -t aci-mcp .

# Run the unified server
docker run --rm -i -e ACI_API_KEY=<ACI_API_KEY> aci-mcp unified-server --linked-account-owner-id <LINKED_ACCOUNT_OWNER_ID>

# Run the apps server
docker run --rm -i -e ACI_API_KEY=<ACI_API_KEY> aci-mcp apps-server --apps <APP1,APP2,...> --linked-account-owner-id <LINKED_ACCOUNT_OWNER_ID>
```

--------------------------------

### Fetch Project State for VibeOps

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Fetches the current state of GitLab, Vercel, and Supabase deployments for an authenticated project. This tool is triggered by an MCP client calling tools/call with name="GET_PROJECT_STATE".

```python
# Triggered by MCP client calling tools/call with name="GET_PROJECT_STATE"
async with httpx.AsyncClient(base_url="https://controller.aci.dev") as client:
    response = await client.get(
        "/v1/projects/self",
        headers={"X-API-KEY": VIBEOPS_API_KEY},
    )
    project_states = response.json()
    # project_states structure:
    # {
    #   "gitlab":  { "resource": { "resource_config": { "name": "my-app", "project_access_token": "glpat-..." } } },
    #   "vercel":  { ... },
    #   "supabase": { ... }
    # }

# The tool returns a detailed prompt to the agent, including:
# - Git remote command:
#   git remote add origin https://vibe:<token>@gitlab.com/vibeops.infra-group/<name>.git
# - Full project state JSON for GitLab, Vercel, Supabase
# - Next.js/TypeScript best practice guidelines (Shadcn UI, Radix, Tailwind, RSC patterns)
# - Supabase Auth redirect URL setup snippet
```

--------------------------------

### Claude Desktop Integration Configuration

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Configure ACI MCP servers in a Claude Desktop `claude_desktop_config.json` for seamless AI agent tool access.

```APIDOC
## Claude Desktop Integration (MCP Client Config)

### Description
Configure ACI MCP servers in a Claude Desktop `claude_desktop_config.json` for seamless AI agent tool access.

### Configuration Example
```json
{
  "mcpServers": {
    "aci-unified": {
      "command": "uvx",
      "args": [
        "aci-mcp",
        "unified-server",
        "--linked-account-owner-id", "your_user_id"
      ],
      "env": {
        "ACI_API_KEY": "your_aci_api_key_here"
      }
    },
    "aci-apps": {
      "command": "uvx",
      "args": [
        "aci-mcp",
        "apps-server",
        "--apps", "BRAVE_SEARCH,GMAIL,GITHUB",
        "--linked-account-owner-id", "your_user_id"
      ],
      "env": {
        "ACI_API_KEY": "your_aci_api_key_here"
      }
    }
  }
}
```
```

--------------------------------

### ACI-MCP CLI Help

Source: https://github.com/aipotheosis-labs/aci-mcp/blob/main/README.md

Displays the help message for the aci-mcp command-line interface, showing available commands and options.

```bash
$ uvx aci-mcp --help
Usage: aci-mcp [OPTIONS] COMMAND [ARGS]...

  Main entry point for the package.

Options:
  --help  Show this message and exit.

Commands:
  apps-server     Start the apps-specific MCP server to access tools...
  unified-server  Start the unified MCP server with unlimited tool access.
```

--------------------------------

### Debug MCP Servers with Inspector

Source: https://github.com/aipotheosis-labs/aci-mcp/blob/main/README.md

Launches the MCP inspector to debug the unified and apps MCP servers. Replace placeholders with your specific configuration.

```bash
# For unified server
npx @modelcontextprotocol/inspector uvx aci-mcp unified-server --linked-account-owner-id <LINKED_ACCOUNT_OWNER_ID>

# For apps server
npx @modelcontextprotocol/inspector uvx aci-mcp apps-server --apps "BRAVE_SEARCH,GMAIL" --linked-account-owner-id <LINKED_ACCOUNT_OWNER_ID>
```

--------------------------------

### Run ACI MCP Apps Server with Docker

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Builds and runs the aci-mcp server using Docker, exposing tools from specified ACI.dev apps. Requires ACI_API_KEY and linked-account-owner-id.

```bash
# Run via Docker
docker build -t aci-mcp .
docker run --rm -i \
  -e ACI_API_KEY=your_aci_api_key_here \
  aci-mcp apps-server \
  --apps "BRAVE_SEARCH,GMAIL" \
  --linked-account-owner-id "user_123"
```

--------------------------------

### Configure MCP Servers for Claude Desktop

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

JSON configuration for Claude Desktop to connect to ACI MCP servers. Specifies command, arguments, and environment variables for each server.

```json
{
  "mcpServers": {
    "aci-unified": {
      "command": "uvx",
      "args": [
        "aci-mcp",
        "unified-server",
        "--linked-account-owner-id", "your_user_id"
      ],
      "env": {
        "ACI_API_KEY": "your_aci_api_key_here"
      }
    },
    "aci-apps": {
      "command": "uvx",
      "args": [
        "aci-mcp",
        "apps-server",
        "--apps", "BRAVE_SEARCH,GMAIL,GITHUB",
        "--linked-account-owner-id", "your_user_id"
      ],
      "env": {
        "ACI_API_KEY": "your_aci_api_key_here"
      }
    }
  }
}
```

--------------------------------

### API Key Validation

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Validates the ACI_API_KEY (or VIBEOPS_API_KEY) at server startup, rejecting placeholder values.

```APIDOC
## API Key Validation — `validate_api_key()`

### Description
Validates the `ACI_API_KEY` (or `VIBEOPS_API_KEY`) at server startup, rejecting placeholder values that users may accidentally copy from documentation examples.

### Method
```python
from aci_mcp.common import validators

# Raises ValueError for missing or placeholder API keys
validators.validate_api_key(os.getenv("ACI_API_KEY"))

# Accepted: any non-empty string that is not a known placeholder
# Rejected with ValueError("Invalid API key"):
validators.validate_api_key(None)
validators.validate_api_key("")
validators.validate_api_key("<YOUR_ACI_API_KEY>")
validators.validate_api_key("ACI_API_KEY")
validators.validate_api_key("<ACI_API_KEY>")
```
```

--------------------------------

### Validate API Key with aci_mcp.common.validators

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Validates the ACI_API_KEY or VIBEOPS_API_KEY environment variable at server startup. Raises ValueError for missing or placeholder keys.

```python
from aci_mcp.common import validators

# Raises ValueError for missing or placeholder API keys
validators.validate_api_key(os.getenv("ACI_API_KEY"))

# Accepted: any non-empty string that is not a known placeholder
# Rejected with ValueError("Invalid API key"):
validators.validate_api_key(None)
validators.validate_api_key("")
validators.validate_api_key("<YOUR_ACI_API_KEY>")
validators.validate_api_key("ACI_API_KEY")
validators.validate_api_key("<ACI_API_KEY>")
```

--------------------------------

### Debug ACI MCP Apps Server with Inspector

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Debugs the apps-server using the MCP Inspector tool. Requires ACI_API_KEY and linked-account-owner-id.

```bash
# Debug with MCP Inspector
npx @modelcontextprotocol/inspector \
  uvx aci-mcp apps-server \
  --apps "BRAVE_SEARCH,GMAIL" \
  --linked-account-owner-id "user_123"
```

--------------------------------

### ACI_SEARCH_FUNCTIONS Tool Usage (Python)

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

This Python code demonstrates how the ACI_SEARCH_FUNCTIONS tool is exposed and internally dispatched. It allows agents to search ACI.dev's function catalog by natural language intent.

```python
# The tool is exposed to MCP clients as:
# {
#   "name": "ACI_SEARCH_FUNCTIONS",
#   "description": "Search for functions available on ACI.dev by intent or keyword.",
#   "inputSchema": {
#     "type": "object",
#     "properties": {
#       "intent": { "type": "string", "description": "Natural language description of what you want to do" },
#       "app_names": { "type": "array", "items": { "type": "string" }, "description": "Optional list of app names to filter by" }
#     },
#     "required": ["intent"]
#   }
# }

# Example agent call (tools/call):
# name: "ACI_SEARCH_FUNCTIONS"
# arguments: { "intent": "send an email" }

# Internally dispatched via:
result = aci.handle_function_call(
    "ACI_SEARCH_FUNCTIONS",
    {"intent": "send an email", "limit": 15, "offset": 0},
    linked_account_owner_id="user_123",
    allowed_apps_only=False,
    format=FunctionDefinitionFormat.ANTHROPIC,
)
# Returns JSON array of matching function definitions the agent can then
# pass to ACI_EXECUTE_FUNCTION.
```

--------------------------------

### Monitor MCP Server Logs

Source: https://github.com/aipotheosis-labs/aci-mcp/blob/main/README.md

Command to continuously display the last 20 lines of MCP server logs, useful for real-time debugging.

```bash
tail -n 20 -f ~/Library/Logs/Claude/mcp*.log
```

--------------------------------

### Search ACI Documentation via REST API

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Queries the ACI.dev documentation using the REST API. This tool is exposed to MCP clients for looking up SDK usage and concepts.

```python
# The tool is exposed to MCP clients as:
# {
#   "name": "ACI_SEARCH_DOCS",
#   "description": "Search for ACI.dev concepts, documentation, Python & TypeScript SDK documentation, and usage examples",
#   "inputSchema": {
#     "type": "object",
#     "properties": {
#       "q": { "type": "string", "description": "The query to search for in the ACI documentation." }
#     },
#     "required": ["q"]
#   }
# }

# Example agent call (tools/call):
# name: "ACI_SEARCH_DOCS"
# arguments: { "q": "how to create a linked account" }

# Internal implementation:
async with httpx.AsyncClient(timeout=30.0) as client:
    response = await client.get(
        "https://api.aci.dev/v1/docs",
        params={"q": "how to create a linked account"},
        headers={"X-API-KEY": ACI_API_KEY},
    )
    # Returns JSON array of relevant documentation snippets.
    # Error cases handled:
    # - Non-200 status  → "Error: API request failed with status <code>."
    # - Timeout         → "Error: Request timed out while querying documentation."
    # - Network error   → "Error: Network error occurred..."
    # - HTTP status err → "Error: HTTP error <code> occurred..."
```

--------------------------------

### Handle Tool Call in MCP Client

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

This Python code handles incoming 'tools/call' requests in an MCP client. It supports an optional override for routing calls to different linked accounts.

```python
import json
import types

# Assume aci and server are imported and configured elsewhere
# LINKED_ACCOUNT_OWNER_ID = "default_owner_id"

@server.call_tool()
async def handle_call_tool(
    name: str,
    arguments: dict
) -> list[types.TextContent | types.ImageContent | types.EmbeddedResource]:

    # Multi-user override: MCP client can pass this extra key to route
    # the call to a different linked account at runtime.
    linked_account_owner_id = LINKED_ACCOUNT_OWNER_ID
    if "aci_override_linked_account_owner_id" in arguments:
        linked_account_owner_id = str(arguments.pop("aci_override_linked_account_owner_id"))

    execution_result = aci.functions.execute(
        function_name=name,                          # e.g. "BRAVE_SEARCH__WEB_SEARCH"
        function_arguments=arguments,                # e.g. {"query": "latest AI news"}
        linked_account_owner_id=linked_account_owner_id,
    )

    if execution_result.success:
        return [types.TextContent(type="text", text=json.dumps(execution_result.data))]
    else:
        return [types.TextContent(type="text", text=f"Failed to execute tool, error: {execution_result.error}")]
```

--------------------------------

### POST /messages/

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

This endpoint is used for posting messages.

```APIDOC
## POST /messages/

### Description
This endpoint is used for posting messages.

### Method
POST

### Endpoint
/messages/
```

--------------------------------

### Run MCP Server via Stdio Transport

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Internal utility to wire an MCP Server instance to a stdio stream. This is the default transport used for local MCP clients like Claude Desktop.

```python
# stdio transport (default) — used for local MCP clients like Claude Desktop
async def run_stdio_async(server: Server):
    async with stdio_server() as (read_stream, write_stream):
        await server.run(read_stream, write_stream, server.create_initialization_options())
```

--------------------------------

### ACI MCP Apps Server: List Tools Handler

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Python code for the `handle_list_tools` function in the apps-server. It fetches functions from ACI.dev in Anthropic format and maps them to MCP tools. Triggered by MCP client's tools/list request.

```python
# Internal handler (src/aci_mcp/apps_server.py)
# Triggered when an MCP client sends a tools/list request.

@server.list_tools() async def handle_list_tools() -> list[types.Tool]:
    functions = aci.functions.search(
        app_names=APPS,               # e.g. ["BRAVE_SEARCH", "GMAIL"]
        allowed_apps_only=False,
        format=FunctionDefinitionFormat.ANTHROPIC,
    )
    return [
        types.Tool(
            name=function["name"],            # e.g. "BRAVE_SEARCH__WEB_SEARCH"
            description=function["description"],
            inputSchema=function["input_schema"],
        )
        for function in functions
    ]
```

--------------------------------

### ACI_SEARCH_FUNCTIONS Tool

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Allows an AI agent to search the full ACI.dev function catalog by natural-language intent or keyword, returning up to 15 matching function definitions.

```APIDOC
## `unified-server` — `ACI_SEARCH_FUNCTIONS` Tool

Allows an AI agent to search the full ACI.dev function catalog by natural-language intent or keyword. Returns up to 15 matching function definitions (limit/offset are set automatically; they are removed from the exposed schema to avoid issues with some MCP clients like Cursor).

```python
# The tool is exposed to MCP clients as:
# {
#   "name": "ACI_SEARCH_FUNCTIONS",
#   "description": "Search for functions available on ACI.dev by intent or keyword.",
#   "inputSchema": {
#     "type": "object",
#     "properties": {
#       "intent": { "type": "string", "description": "Natural language description of what you want to do" },
#       "app_names": { "type": "array", "items": { "type": "string" }, "description": "Optional list of app names to filter by" }
#     },
#     "required": ["intent"]
#   }
# }

# Example agent call (tools/call):
# name: "ACI_SEARCH_FUNCTIONS"
# arguments: { "intent": "send an email" }

# Internally dispatched via:
result = aci.handle_function_call(
    "ACI_SEARCH_FUNCTIONS",
    {"intent": "send an email", "limit": 15, "offset": 0},
    linked_account_owner_id="user_123",
    allowed_apps_only=False,
    format=FunctionDefinitionFormat.ANTHROPIC,
)
# Returns JSON array of matching function definitions the agent can then
# pass to ACI_EXECUTE_FUNCTION.
```
```

--------------------------------

### ACI_EXECUTE_FUNCTION Tool

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Executes any function on ACI.dev by name with supplied arguments, intended to be called after ACI_SEARCH_FUNCTIONS. Supports multi-user routing via `aci_override_linked_account_owner_id`.

```APIDOC
## `unified-server` — `ACI_EXECUTE_FUNCTION` Tool

Executes any function on ACI.dev by name with supplied arguments. Intended to be called after `ACI_SEARCH_FUNCTIONS` identifies the right function. Also supports the `aci_override_linked_account_owner_id` key for multi-user routing.

```python
# The tool is exposed to MCP clients as:
# {
#   "name": "ACI_EXECUTE_FUNCTION",
#   "description": "Execute a function on ACI.dev by name.",
#   "inputSchema": {
#     "type": "object",
#     "properties": {
#       "function_name": { "type": "string" },
#       "function_arguments": { "type": "object" }
#     },
#     "required": ["function_name", "function_arguments"]
#   }
# }

# Example agent call (tools/call):
# name: "ACI_EXECUTE_FUNCTION"
# arguments: {
#   "function_name": "GMAIL__SEND_EMAIL",
#   "function_arguments": {
#     "to": "alice@example.com",
#     "subject": "Hello from ACI",
#     "body": "This email was sent by an AI agent via ACI MCP."
#   },
#   "aci_override_linked_account_owner_id": "other_user_456"  # optional multi-user override
# }
```
```

--------------------------------

### Run MCP Server via SSE Transport

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Internal utility to wire an MCP Server instance to an SSE/HTTP endpoint powered by Starlette and Uvicorn. This is used for remote/network MCP clients.

```python
# SSE transport — used for remote/network MCP clients
# Exposes:
#   GET  /sse        → SSE connection endpoint
```

--------------------------------

### ACI MCP Apps Server: Call Tool Handler

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Placeholder for the `handle_call_tool` function in the apps-server. This function executes a named tool when invoked by an MCP client, delegating to `aci.functions.execute()` and returning the result.

```python
# Internal handler (src/aci_mcp/apps_server.py)

```

--------------------------------

### Execute ACI Function Call

Source: https://context7.com/aipotheosis-labs/aci-mcp/llms.txt

Internally dispatches function calls to the ACI. Use this for executing predefined ACI functions like sending emails.

```python
result = aci.handle_function_call(
    "ACI_EXECUTE_FUNCTION",
    {
        "function_name": "GMAIL__SEND_EMAIL",
        "function_arguments": { "to": "alice@example.com", "subject": "Hello", "body": "..." }
    },
    linked_account_owner_id="user_123",
    allowed_apps_only=False,
    format=FunctionDefinitionFormat.ANTHROPIC,
)
# Returns JSON-serialized execution result.
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.