Pydantic AI

A decorator that puts a permission check in front of a Pydantic AI tool, carrying the agent identity through RunContext so the audit event knows who ran it.

Pydantic AI Integration

Type-safe agent identity for Pydantic AI agents. Uses dependency injection to provide identity, permission checking, and audit logging through RunContext.

Installation

pip install vorim pydantic-ai

bash

VorimDeps Dependency

Pass VorimDeps as your agent's dependency type. It provides check(), emit(), grant(), and verify() methods on the context.

from vorim.pydantic_ai import VorimDeps
from pydantic_ai import Agent, RunContext

agent = Agent('openai:gpt-4o', deps_type=VorimDeps)

@agent.tool
async def fetch_data(ctx: RunContext[VorimDeps], query: str) -> str:
    # Check permission before acting
    check = ctx.deps.check('agent:read')
    if not check.get('allowed'):
        return 'Permission denied: agent:read not granted'

    result = do_something(query)

    # Log the action
    ctx.deps.emit('data.fetch', outcome='success')
    return result

# Run with Vorim identity
deps = VorimDeps(api_key='agid_sk_...', agent_id='agid_abc123')
result = await agent.run('Fetch the latest data', deps=deps)

typescript

vorim_tool Decorator

Automatically checks permissions and logs audit events on every tool call. No manual check()/emit() needed.

from vorim.pydantic_ai import VorimDeps, vorim_tool

agent = Agent('openai:gpt-4o', deps_type=VorimDeps)

@agent.tool
@vorim_tool(scope='agent:read', action='data.fetch')
async def fetch_data(ctx: RunContext[VorimDeps], query: str) -> str:
    # Permission checked automatically before this runs
    # Audit event emitted automatically after this returns
    return do_something(query)

# If permission is denied, tool returns 'Permission denied'
# and a denial audit event is logged

typescript

Quick Start with create_vorim_agent

from vorim.pydantic_ai import create_vorim_agent

agent, deps = create_vorim_agent(
    model='openai:gpt-4o',
    api_key='agid_sk_...',
    agent_name='research-agent',
    scopes=['agent:read', 'agent:execute'],
    system_prompt='You are a research agent.',
)

result = await agent.run('Find the latest papers', deps=deps)

typescript

Need Help?

Book a demo for a walkthrough, or contact us for support. For enterprise needs, reach out at sales@vorim.ai.