Reins

Take Control of Your AI Agents

The runtime platform that makes AI agents debuggable, affordable, and reliable.
Existing tools show you what happened. Reins lets you control what happens next.

Quick Start • Features • Frameworks • Docs • Why Reins

pip install reins

from reins import trace

@trace(budget="$0.50", on_exceed="degrade")
async def my_agent(task: str):
    response = await client.messages.create(model="claude-sonnet-4-20250514", ...)
    return response
    # When budget runs low → auto-switches to claude-haiku (not crash)

One decorator. Budget control + cost tracking + auto-degradation + trace recording.

Why Reins?

Enterprise LLM Spending is Exploding — But Cost Controls Haven't Kept Up

Enterprise LLM API spending has surged from $1.8B (2023 H2) to $8.4B (2025 H1) — a 4.7x increase in 18 months. Menlo Ventures projects it will hit $15B by 2026 if current velocity holds. Total enterprise GenAI investment reached $37B in 2025, tripling from $11.5B in 2024.

Yet the tools to govern this spending are shockingly primitive:

Capability	Available Today?
Cost tracking (after the fact)	Widely available (Langfuse, LangSmith, Helicone)
Hard budget limits (block when exceeded)	Partial (LiteLLM, Portkey)
Auto-degradation (switch to cheaper model)	No existing tool
Circuit breakers (halt runaway loops)	No existing tool
Per-agent budget scoping	No existing tool

Real Cost Overruns Are Already Happening

$47K LangChain Loop (Nov 2025): Four agents in a research pipeline entered an infinite conversation loop for 11 days. The team assumed growing costs were "organic growth" — until the $47,000 bill arrived.
$47K Retry Storm (Feb 2026): A data enrichment agent misinterpreted an API error code, running 2.3 million API calls over a weekend. Only the external API's rate limiter slowed it down — not the team's own controls.
Gartner (2025): Over 40% of agentic AI projects will be canceled by 2027 due to escalating costs, unclear value, or inadequate risk controls.

Agent Reliability is a Crisis

Top SWE-bench Verified score: 79% — but real-world performance overestimates by up to 54%
Agent success rates decline exponentially with task duration. Claude Sonnet's "half-life" is ~59 minutes
A survey of 306 practitioners found reliability is the #1 barrier to enterprise agent adoption

Who Needs Cost Governance?

Industry	AI Spend (2025)	Growth	Key Concern
Healthcare	~$1.5B in vertical AI	3.3x YoY	Compliance + cost predictability
Financial Services	23.7% of enterprise AI market	Steady	Risk controls + audit trails
Legal	$650M market	Fast-growing	Per-case cost attribution
Customer Service	Largest agent deployment sector	Rapid	Per-conversation cost caps

86% of enterprises plan to increase AI budgets in 2026 (Deloitte). The question isn't whether to spend — it's whether to spend blindly.

Quick Start

For Your Own Agents

pip install reins

from reins import trace

@trace(budget="$0.50", on_exceed="degrade")
async def my_agent(task: str):
    client = anthropic.AsyncAnthropic()
    response = await client.messages.create(
        model="claude-sonnet-4-20250514",
        max_tokens=1024,
        messages=[{"role": "user", "content": task}],
    )
    return response.content[0].text

When the budget runs low, Reins automatically switches claude-sonnet to claude-haiku — your agent keeps running, just cheaper.

For Claude Code

pip install reins[proxy]

# Start local proxy with $5/day budget
reins proxy --port 8082 --budget '$5/day' --on-exceed degrade

# Point Claude Code at it
export ANTHROPIC_BASE_URL=http://localhost:8082

# Use Claude Code normally — Reins controls costs transparently
claude "refactor this module"

Check Your Costs

# Cost report
reins report

# Trace visualization
reins trace list
reins trace show <run_id>

# Context health analysis
reins health <run_id>

# Step-by-step replay
reins replay <run_id>

Features

Modular Architecture — Use What You Need

pip install reins              # Core: auto-tracing + DuckDB storage
pip install reins[budget]      # + Cost governance
pip install reins[lens]        # + Debugging (replay, context health)
pip install reins[pulse]       # + Reliability (guardrails, evaluation)
pip install reins[all]         # Everything

Core (Always Installed)

Auto-instrumentation: Monkey-patches Anthropic & OpenAI SDKs — zero code changes to capture all LLM calls
DuckDB storage: Embedded, zero-config. No Redis, no Postgres, no Docker
SQL queries: reins query "SELECT * FROM spans WHERE cost > 0.1"
Streaming support: Transparent interception of streaming responses

Budget Module

Per-run budgets: @trace(budget="$0.50") — hard cap per agent execution
4 exceed strategies: degrade (auto-switch model) / pause / alert / reject
Model degradation chains: opus → sonnet → haiku, o3 → gpt-4o → gpt-4o-mini
Circuit breaker: Auto-halts runaway agent loops (>30 calls/minute)
Budget persistence: Survives process restarts, daily/monthly auto-reset
Cost anomaly detection: Alerts when a run costs 3x the historical average
YAML team budgets: Organization → team → agent hierarchy

# reins.yaml
budgets:
  daily: $10.00
  agents:
    research_agent: { per_run: $2.00, on_exceed: degrade }
    code_agent: { per_run: $0.50, on_exceed: reject }

Trace Module

reins trace list: Recent runs table with cost, status, degradation count
reins trace show: Colored terminal call tree (LLM + tool calls)
reins trace export --format otel: OTLP JSON export for Grafana/Datadog
Cross-agent correlation: Automatic trace_id propagation

Lens Module

reins replay: Interactive step-by-step agent replay (Enter/auto-play/quit)
reins health: Context health curve with ASCII chart + per-span breakdown
Context Rot detection: 3-metric composite score (utilization, efficiency, duplication)
Root cause analysis: Automatic causal chain for failures

Pulse Module (Phase 3)

Runtime evaluators (coherence, instruction following)
Guardrail engine (PII detection, SQL injection prevention)
Automatic regression test generation from failed runs

Proxy Mode

Transparent reverse proxy for Claude Code, aider, Cursor, or any tool that respects ANTHROPIC_BASE_URL
Full budget enforcement at the HTTP layer
Zero changes to the upstream tool

Framework Support

Reins integrates with 10+ agent frameworks through 4 adapters:

Adapter	Frameworks	Integration
`ReinsCallbackHandler`	LangChain, LangGraph, LangFlow	`ChatAnthropic(callbacks=[handler])`
`ReinsTracingProcessor`	OpenAI Agents SDK	`add_trace_processor(processor)`
`instrument_crew()`	CrewAI	`instrument_crew(crew)`
`ReinsSpanExporter`	Semantic Kernel, Pydantic AI, Haystack, Mastra	OTel SpanProcessor

LangChain / LangGraph

from reins.adapters.langchain import ReinsCallbackHandler

handler = ReinsCallbackHandler(agent_name="my_langchain_agent")
llm = ChatAnthropic(model="claude-sonnet-4-20250514", callbacks=[handler])
chain = prompt | llm | parser
result = chain.invoke({"input": "..."})

OpenAI Agents SDK

from reins.adapters.openai_agents import ReinsTracingProcessor
from agents import Agent, Runner, add_trace_processor

add_trace_processor(ReinsTracingProcessor())

agent = Agent(name="assistant", model="gpt-4o")
result = Runner.run_sync(agent, "Hello!")

CrewAI

from reins.adapters.crewai import instrument_crew
from crewai import Crew, Agent, Task

crew = Crew(agents=[...], tasks=[...])
instrument_crew(crew)  # One line — instruments all agents and tasks
result = crew.kickoff()

Any OTel-Native Framework (Semantic Kernel, Pydantic AI, Haystack, Mastra)

from reins.adapters.otel import ReinsSpanProcessor
from opentelemetry.sdk.trace import TracerProvider

provider = TracerProvider()
provider.add_span_processor(ReinsSpanProcessor())
# Now any OTel-instrumented framework is automatically traced by Reins

How It Compares

	Reins	LiteLLM	Langfuse	Helicone	Portkey
Architecture	SDK + Proxy	Proxy only	SDK	Proxy	Proxy/SDK
Infrastructure	Zero (embedded DuckDB)	Redis + Postgres	PostgreSQL	Cloud	Cloud
Budget enforcement	Smart degradation	Hard reject (400)	None	Rate limit	Key limit
Auto model switch	Yes	No	No	No	No
Circuit breaker	Yes	No	No	No	No
Per-agent budgets	Yes	Per-key	No	No	Partial
Agent replay	Yes	No	No	No	No
Context health	Yes	No	No	No	No
Framework adapters	10+ frameworks	N/A	12+	N/A	5+
Setup	`pip install reins`	`docker-compose up`	`docker-compose up`	Cloud signup	Cloud signup
Open source	BSL 1.1 (→ Apache 2030)	Enterprise paywall	MIT	MIT	Proprietary

One-line difference: LiteLLM is an API Gateway (needs infra, hard-rejects on exceed). Reins is an Agent Runtime (zero-infra, degrades gracefully).

Architecture

Modules communicate via an event bus — install only what you need, they auto-cooperate when co-installed. For example: Lens detects context rot → notifies Budget → Budget reduces remaining allocation.

Documentation

Product Requirements (PRD) — What we build and why
Technical Design — Architecture, data models, API design

Market Context

Metric	Value	Source
Enterprise LLM API spend (2025 H1)	$8.4B	Menlo Ventures
YoY GenAI enterprise investment growth	3.2x ($11.5B → $37B)	Menlo Ventures
Projected LLM API spend (2026)	$15B+	Menlo Ventures
Agentic AI projects to be canceled by 2027	>40%	Gartner
Agent reliability as #1 enterprise barrier	72% of practitioners	Pan et al. (2025)
Enterprises increasing AI budget in 2026	86%	Deloitte State of AI 2026
Langfuse: SDK installs/month	26M+	ClickHouse acquisition
LangChain valuation (Oct 2025)	$1.25B	Series B
Healthcare vertical AI spend (2025)	$1.5B (3.3x YoY)	Menlo Ventures

Development

# Clone
git clone https://github.com/catyans/reins.git
cd reins

# Install with dev deps
pip install -e ".[dev,all]"

# Run tests
pytest tests/ -v

# Generate charts
python scripts/generate_charts.py

99 tests across unit and integration suites.

License

Business Source License 1.1 (BSL 1.1)

Free for: personal use, internal enterprise use, academic research, contributing back
Not allowed: building a competing commercial AI agent cost governance product/service
Auto-converts to Apache 2.0 on April 4, 2030

For commercial licensing inquiries: 237344440@qq.com

Author

Yanshu Wang (@catyans) — https://catyans.github.io

Reins: Take control of your AI agents.
pip install reins

Name		Name	Last commit message	Last commit date
Latest commit History 16 Commits
.github		.github
docs		docs
examples		examples
scripts		scripts
src/reins		src/reins
tests		tests
.gitignore		.gitignore
CHANGELOG.md		CHANGELOG.md
CONTRIBUTING.md		CONTRIBUTING.md
LICENSE		LICENSE
README.md		README.md
SECURITY.md		SECURITY.md
pyproject.toml		pyproject.toml

Folders and files

Latest commit

History

Repository files navigation

Reins

Why Reins?

Enterprise LLM Spending is Exploding — But Cost Controls Haven't Kept Up

Real Cost Overruns Are Already Happening

Agent Reliability is a Crisis

Who Needs Cost Governance?

Quick Start

For Your Own Agents

For Claude Code

Check Your Costs

Features

Modular Architecture — Use What You Need

Core (Always Installed)

Budget Module

Trace Module

Lens Module

Pulse Module (Phase 3)

Proxy Mode

Framework Support

LangChain / LangGraph

OpenAI Agents SDK

CrewAI

Any OTel-Native Framework (Semantic Kernel, Pydantic AI, Haystack, Mastra)

How It Compares

Architecture

Documentation

Market Context

Development

License

Author

About

Resources

License

Contributing

Security policy

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages