9.8 KiB
Building Tools for Aden
This guide explains how to create new tools for the Aden agent framework using FastMCP.
Quick Start Checklist
- Create folder under
src/aden_tools/tools/<tool_name>/ - Implement a
register_tools(mcp: FastMCP)function using the@mcp.tool()decorator - Add a
README.mddocumenting your tool - Register in
src/aden_tools/tools/__init__.py - Add tests in
tests/tools/
Tool Structure
Each tool lives in its own folder:
src/aden_tools/tools/my_tool/
├── __init__.py # Export register_tools function
├── my_tool.py # Tool implementation
└── README.md # Documentation
Implementation Pattern
Tools use FastMCP's native decorator pattern:
from fastmcp import FastMCP
def register_tools(mcp: FastMCP) -> None:
"""Register my tools with the MCP server."""
@mcp.tool()
def my_tool(
query: str,
limit: int = 10,
) -> dict:
"""
Search for items matching a query.
Use this when you need to find specific information.
Args:
query: The search query (1-500 chars)
limit: Maximum number of results (1-100)
Returns:
Dict with search results or error dict
"""
# Validate inputs
if not query or len(query) > 500:
return {"error": "Query must be 1-500 characters"}
if limit < 1 or limit > 100:
limit = max(1, min(100, limit))
try:
# Your implementation here
results = do_search(query, limit)
return {
"query": query,
"results": results,
"total": len(results),
}
except Exception as e:
return {"error": f"Search failed: {str(e)}"}
Exporting the Tool
In src/aden_tools/tools/my_tool/__init__.py:
from .my_tool import register_tools
__all__ = ["register_tools"]
In src/aden_tools/tools/__init__.py, add to _TOOL_MODULES:
_TOOL_MODULES = [
# ... existing tools
"my_tool",
]
Credential Management
Tools fall into two categories based on whether they need external API credentials:
| Signature | Meaning | CI Enforcement |
|---|---|---|
register_tools(mcp) |
No credentials needed | ✅ Just works |
register_tools(mcp, credentials=None) |
Requires credentials | ⚠️ Must have CredentialSpec |
This is enforced by CI — if your register_tools accepts a credentials parameter, every tool it registers must appear in a CredentialSpec.tools list. Otherwise, CI will fail with a clear error message.
Tools WITHOUT Credentials (Simple Case)
If your tool doesn't need external API keys (file operations, local processing, etc.), just use the simple signature:
def register_tools(mcp: FastMCP) -> None:
"""Register tools that don't need credentials."""
@mcp.tool()
def my_local_tool(path: str) -> dict:
"""Process a local file."""
# No credentials needed - just do the work
return {"result": process_file(path)}
That's it! No additional configuration needed.
Tools WITH Credentials (Integration Case)
For tools requiring API keys, follow these steps:
Step 1: Add the credentials parameter
from typing import TYPE_CHECKING
if TYPE_CHECKING:
from aden_tools.credentials import CredentialStoreAdapter
def register_tools(
mcp: FastMCP,
credentials: CredentialStoreAdapter | None = None,
) -> None:
@mcp.tool()
def my_api_tool(query: str) -> dict:
"""Tool that requires an API key."""
# Use credentials adapter if provided, fallback to direct env access
if credentials is not None:
api_key = credentials.get("my_api")
else:
api_key = os.getenv("MY_API_KEY")
if not api_key:
return {
"error": "MY_API_KEY environment variable not set",
"help": "Get an API key at https://example.com/api-keys",
}
# Use the API key...
Step 2: Create a CredentialSpec
Find the appropriate category file in src/aden_tools/credentials/ or create a new one:
| Category | File | Examples |
|---|---|---|
| LLM providers | llm.py |
anthropic, openai |
| Search tools | search.py |
brave_search, google_search |
| Email providers | email.py |
resend, google/gmail |
| GitHub | github.py |
github |
| CRM | hubspot.py |
hubspot |
| Messaging | slack.py |
slack |
Add your credential spec:
# In credentials/<category>.py
from .base import CredentialSpec
MY_CREDENTIALS = {
"my_api": CredentialSpec(
env_var="MY_API_KEY",
tools=["my_api_tool"], # IMPORTANT: List ALL tool names this credential covers
required=True,
help_url="https://example.com/api-keys",
description="API key for My Service",
# Credential store mapping
credential_id="my_api",
credential_key="api_key",
),
}
Important: The tools list must include every tool name that your register_tools function creates. CI will fail if any tool is missing.
Step 3: Merge into CREDENTIAL_SPECS
If you created a new category file, import and merge it in credentials/__init__.py:
from .my_category import MY_CREDENTIALS
CREDENTIAL_SPECS = {
**LLM_CREDENTIALS,
**SEARCH_CREDENTIALS,
**MY_CREDENTIALS, # Add new category
}
__all__ = [
# ... existing exports
"MY_CREDENTIALS",
]
Step 4: Update register_all_tools
In tools/__init__.py, add your tool registration with credentials:
from .my_tool import register_tools as register_my_tool
def register_all_tools(mcp: FastMCP, credentials=None) -> list[str]:
# ... existing registrations
# Tools that need credentials
register_my_tool(mcp, credentials=credentials)
return [
# ... existing tool names
"my_api_tool",
]
CI Enforcement Rules
The following conformance tests run in CI (tests/integrations/test_spec_conformance.py):
| Test | What It Checks |
|---|---|
TestModuleStructure |
Every tool module exports register_tools |
TestRegisterToolsSignature |
Correct function signature (mcp param, optional credentials) |
TestCredentialSpecFields |
All CredentialSpec fields are complete (env_var, help_url, description, credential_id, credential_key) |
TestSpecToolsMatchRegistered |
Tool names in spec.tools actually exist |
TestCredentialCoverage |
Every tool from a module with credentials param has a spec |
If TestCredentialCoverage fails, you'll see:
Tool 'my_new_tool' from module 'my_tool' accepts credentials but has no CredentialSpec.
Fix by either:
1. Adding a CredentialSpec in credentials/<category>.py with tools=['my_new_tool'], or
2. Removing 'credentials' param from register_tools() if this tool doesn't need credentials
Testing with Mock Credentials
from aden_tools.credentials import CredentialStoreAdapter
def test_my_tool_with_valid_key(mcp):
creds = CredentialStoreAdapter.for_testing({"my_api": "test-key"})
register_tools(mcp, credentials=creds)
tool_fn = mcp._tool_manager._tools["my_api_tool"].fn
result = tool_fn(query="test")
# Assertions...
When Validation Happens
Credentials are validated when an agent is loaded (via AgentRunner.validate()), not at MCP server startup. This means:
- The MCP server always starts (even if credentials are missing)
- When you load an agent, validation checks which tools it needs
- If credentials are missing, you get a clear error:
Cannot run agent: Missing credentials
The following tools require credentials that are not set:
web_search requires BRAVE_SEARCH_API_KEY
API key for Brave Search
Get an API key at: https://brave.com/search/api/
Set via: export BRAVE_SEARCH_API_KEY=your_key
Set these environment variables and re-run the agent.
Best Practices
Error Handling
Return error dicts instead of raising exceptions:
@mcp.tool()
def my_tool(**kwargs) -> dict:
try:
result = do_work()
return {"success": True, "data": result}
except SpecificError as e:
return {"error": f"Failed to process: {str(e)}"}
except Exception as e:
return {"error": f"Unexpected error: {str(e)}"}
Return Values
- Return dicts for structured data
- Include relevant metadata (query, total count, etc.)
- Use
{"error": "message"}for errors
Documentation
The docstring becomes the tool description in MCP. Include:
- What the tool does
- When to use it
- Args with types and constraints
- What it returns
Every tool folder needs a README.md with:
- Description and use cases
- Usage examples
- Argument table
- Environment variables (if any)
- Error handling notes
Testing
Place tests in tests/tools/test_{{tool_name}}.py:
import pytest
from fastmcp import FastMCP
from aden_tools.tools.{{tool_name}} import register_tools
@pytest.fixture
def mcp():
"""Create a FastMCP instance with tools registered."""
server = FastMCP("test")
register_tools(server)
return server
def test_my_tool_basic(mcp):
"""Test basic tool functionality."""
tool_fn = mcp._tool_manager._tools["my_tool"].fn
result = tool_fn(query="test")
assert "results" in result
def test_my_tool_validation(mcp):
"""Test input validation."""
tool_fn = mcp._tool_manager._tools["my_tool"].fn
result = tool_fn(query="")
assert "error" in result
Mock external APIs to keep tests fast and deterministic.
Naming Conventions
- Folder name:
snake_casewith_toolsuffix (e.g.,file_read_tool) - Function name:
snake_case(e.g.,file_read) - Tool description: Clear, actionable docstring