CrewAI + Govyn — Govern Your CrewAI Agents

CrewAI orchestrates multiple agents that collaborate on tasks — and each agent makes its own LLM calls. A single crew run can trigger dozens of completions across multiple agents, making cost control nearly impossible without an external governance layer.

How it works

CrewAI
Your agents
─ HTTPS →
Govyn Proxy
Policy · Budget · Logs
─ API →
OpenAI API
LLM provider

Step-by-step setup

1

Start the Govyn proxy

bash 
npx govyn start --config govyn.yaml
2

Configure CrewAI agents to use Govyn

python 
from crewai import Agent, Crew, Task
from langchain_openai import ChatOpenAI

llm = ChatOpenAI(
    model="gpt-4o",
    openai_api_base="http://localhost:4111/v1",
    openai_api_key="gvn_agent_crew_researcher"
)

researcher = Agent(
    role="Researcher",
    goal="Find relevant data",
    llm=llm
)
3

Run your crew with governance

python 
task = Task(
    description="Research Q4 market trends",
    agent=researcher
)

crew = Crew(agents=[researcher], tasks=[task])
result = crew.kickoff()

Example policy

Define governance rules for your CrewAI agents in a simple YAML file.

govyn.yaml 
agents:
  crew_researcher:
    budget:
      daily: $3.00
      per_task: $1.00
    models:
      allow: [gpt-4o-mini]
    rate_limit:
      requests_per_minute: 20
  crew_writer:
    budget:
      daily: $5.00
      per_task: $2.00
    models:
      allow: [gpt-4o]
    rate_limit:
      requests_per_minute: 15

Why use Govyn with CrewAI?

Independent budgets for each agent in a crew
Per-task spending limits
Prevent expensive model usage by role
Rate limiting to avoid agent feedback loops
Full replay of multi-agent conversations
Monitor which agents consume the most budget

Get started in 5 minutes

Add governance to your CrewAI agents with a single config change. No code rewrites.

Read the docs

Frequently asked questions

Can I give each CrewAI agent its own budget?
Yes. Each agent gets a unique Govyn API key with independent budget, model, and rate limit policies. Your researcher agent can have a $3/day limit on gpt-4o-mini while your writer agent gets $5/day on gpt-4o.
How does Govyn handle CrewAI's agent delegation?
When one CrewAI agent delegates to another, each agent's LLM calls go through Govyn with their respective API keys. Delegation works transparently — Govyn enforces each agent's policies independently regardless of the orchestration pattern.
Can I see which agent in my crew is spending the most?
Yes. Govyn's dashboard and replay logs are keyed by agent API key, so you can see exactly how much each crew member is spending, which models they're using, and how many requests they're making.

Related integrations

CrewAI + Anthropic

Govern CrewAI multi-agent crews using Claude. Set per-agent budgets, enforce model policies, and replay every conversation.
LangChain + OpenAI

Add budget limits, policy enforcement, and full replay to LangChain agents using OpenAI. Five-minute setup, zero code changes.
OpenAI Agents SDK + OpenAI

Add governance to OpenAI Agents SDK. Enforce budgets, model policies, and audit trails for every agent run. Five-minute setup.

Explore more

FROM OUR BLOG
Proxy vs SDK: Why Architecture Matters for AI Agent Governance

SDK wrappers are door locks. Proxies are walls. A deep technical comparison of both governance architectures for AI agents in production.

FROM OUR BLOG
We Cut Our AI API Bill by 73% Without Changing a Single Line of Agent Code

How smart model routing through a proxy cut our OpenAI and Anthropic bill from $2,140/mo to $578/mo. Zero code changes. Just YAML.

POLICY TEMPLATE
Budget Control Policy

Set daily and monthly spending limits for AI agents. Prevent runaway costs with hard budget caps enforced at the proxy level.

POLICY TEMPLATE
Loop Detection Policy

Detect and stop AI agent infinite loops automatically. Prevent runaway tool calls and recursive chains with proxy-level loop detection.