CaveAgent: Transforming LLMs into Stateful Runtime Operators

"From text-in-text-out to (text&object)-in-(text&object)-out"

---

Most LLM agents operate under a text-in-text-out paradigm, with tool interactions constrained to JSON primitives. CaveAgent breaks this with Stateful Runtime Management—a persistent Python runtime with direct variable injection and retrieval:

• Inject any Python object into the runtime—DataFrames, models, database connections, custom class instances—as first-class variables the LLM can manipulate
• Persist state across turns without serialization; objects live in the runtime, not in the context window
• Retrieve manipulated objects back as native Python types for downstream

Table of Contents

• Installation
• Hello World
• Examples
  • Function Calling
  • Stateful Object Interactions
  • Multi-Agent Coordination
  • Real-time Streaming
  • Security Rules
• Agent Skills
  • Creating a Skill
  • How Skills Load
  • Using Skills
  • Injection Module
• Features
• Configuration
• LLM Provider Support

Installation

pip install 'cave-agent[all]'

Choose your installation:

# OpenAI support
pip install 'cave-agent[openai]'

# 100+ LLM providers via LiteLLM
pip install 'cave-agent[litellm]'

Hello World

import asyncio
from cave_agent import CaveAgent
from cave_agent.runtime import PythonRuntime, Variable, Function
from cave_agent.models import LiteLLMModel

model = LiteLLMModel(model_id="model-id", api_key="your-api-key", custom_llm_provider="openai")

async def main():
    def reverse(s: str) -> str:
        """Reverse a string"""
        return s[::-1]

    runtime = PythonRuntime(
        variables=[
            Variable("secret", "!dlrow ,olleH", "A reversed message"),
            Variable("greeting", "", "Store the reversed message"),
        ],
        functions=[Function(reverse)],
    )
    agent = CaveAgent(model, runtime=runtime)
    response = await agent.run("Reverse the secret")
    print(runtime.retrieve("greeting"))  # Hello, world!
    print(response.content)              # Agent's text response

asyncio.run(main())

Examples

Function Calling

# Inject functions and variables into runtime
runtime = PythonRuntime(
    variables=[Variable("tasks", [], "User's task list")],
    functions=[Function(add_task), Function(complete_task)],
)
agent = CaveAgent(model, runtime=runtime)

await agent.run("Add 'buy groceries' to my tasks")
print(runtime.retrieve("tasks"))  # [{'name': 'buy groceries', 'done': False}]

See examples/basic_usage.py for a complete example.

Stateful Object Interactions

# Inject objects with methods - LLM can call them directly
runtime = PythonRuntime(
    types=[Type(Light), Type(Thermostat)],
    variables=[
        Variable("light", Light("Living Room"), "Smart light"),
        Variable("thermostat", Thermostat(), "Home thermostat"),
    ],
)
agent = CaveAgent(model, runtime=runtime)

await agent.run("Dim the light to 20% and set thermostat to 22°C")
light = runtime.retrieve("light")  # Object with updated state

See examples/object_methods.py for a complete example.

Multi-Agent Coordination

# Sub-agents with their own runtimes
cleaner_agent = CaveAgent(model, runtime=PythonRuntime(variables=[
    Variable("data", [], "Input"), Variable("cleaned_data", [], "Output"),
]))

analyzer_agent = CaveAgent(model, runtime=PythonRuntime(variables=[
    Variable("data", [], "Input"), Variable("insights", {}, "Output"),
]))

# Orchestrator controls sub-agents as first-class objects
orchestrator = CaveAgent(model, runtime=PythonRuntime(variables=[
    Variable("raw_data", raw_data, "Raw dataset"),
    Variable("cleaner", cleaner_agent, "Cleaner agent"),
    Variable("analyzer", analyzer_agent, "Analyzer agent"),
]))

# Inject → trigger → retrieve
await orchestrator.run("Clean raw_data using cleaner, then analyze using analyzer")
insights = analyzer.runtime.retrieve("insights")

See examples/multi_agent.py for a complete example.

Real-time Streaming

async for event in agent.stream_events("Analyze this data"):
    if event.type.value == 'code':
        print(f"Executing: {event.content}")
    elif event.type.value == 'execution_output':
        print(f"Result: {event.content}")

See examples/stream.py for a complete example.

Security Rules

# Block dangerous operations with AST-based validation
rules = [
    ImportRule({"os", "subprocess", "sys"}),
    FunctionRule({"eval", "exec", "open"}),
    AttributeRule({"__globals__", "__builtins__"}),
]
runtime = PythonRuntime(security_checker=SecurityChecker(rules))

More Examples

• Basic Usage: Function calling and object processing
• Runtime State: State management across interactions
• Object Methods: Class methods and complex objects
• Multi-Turn: Conversations with state persistence
• Multi-Agent: Data pipeline with multiple agents
• Stream: Streaming responses and events

Agent Skills

CaveAgent implements the Agent Skills open standard—a portable format for packaging instructions that agents can discover and use. Originally developed by Anthropic and now supported across the AI ecosystem (Claude, Gemini CLI, Cursor, VS Code, and more), Skills enable agents to acquire domain expertise on-demand.

Creating a Skill

A Skill is a directory containing a SKILL.md file with YAML frontmatter:

my-skill/
├── SKILL.md           # Required: Skill definition and instructions
└── injection.py       # Optional: Functions/variables/types to inject (CaveAgent extension)

SKILL.md structure:

---
name: data-processor
description: Process and analyze datasets with statistical methods. Use when working with data analysis tasks.
---

# Data Processing Instructions

## Quick Start
Use the injected functions to analyze datasets...

## Workflows
1. Activate the skill to load injected functions
2. Apply statistical analysis using the provided functions
3. Return structured results

Required fields: name (max 64 chars, lowercase with hyphens) and description (max 1024 chars)

Optional fields: license, compatibility, metadata

How Skills Load (Progressive Disclosure)

Skills use progressive disclosure to minimize context usage:

Level	When Loaded	Content
Metadata	At startup	name and description from YAML frontmatter (~100 tokens)
Instructions	When activated	SKILL.md body with guidance (loaded on-demand)

Using Skills

from cave_agent import CaveAgent
from cave_agent.skills import SkillDiscovery

# Load skills from directory
agent = CaveAgent(model=model, skills_dir="./skills")

# Or load specific skills
skill = SkillDiscovery.from_file("./my-skill/SKILL.md")
agent = CaveAgent(model=model, skills=[skill])

When skills are loaded, the agent gains access to the activate_skill(skill_name) runtime function to activate a skill and load its instructions.

Injection Module (CaveAgent Extension)

CaveAgent extends the Agent Skills standard with injection.py—allowing skills to inject functions, variables, and types directly into the runtime when activated:

from cave_agent.runtime import Function, Variable, Type
from dataclasses import dataclass

def analyze_data(data: list) -> dict:
    """Analyze data and return statistics."""
    return {"mean": sum(data) / len(data), "count": len(data)}

@dataclass
class AnalysisResult:
    mean: float
    count: int
    status: str

CONFIG = {"threshold": 0.5, "max_items": 1000}

__exports__ = [
    Function(analyze_data, description="Analyze data statistically"),
    Variable("CONFIG", value=CONFIG, description="Analysis configuration"),
    Type(AnalysisResult, description="Result structure"),
]

When activate_skill() is called, these exports are automatically injected into the runtime namespace.

See examples/skill_data_processor.py for a complete example.

Features

• Code-Based Function Calling: Leverages LLM's natural coding abilities instead of rigid JSON schemas
• Secure Runtime Environment:
  • Inject Python objects, variables, and functions as tools
  • Rule-based security validation prevents dangerous code execution
  • Flexible security rules: ImportRule, FunctionRule, AttributeRule, RegexRule
  • Customizable security policies for different use cases
  • Access execution results and maintain state across interactions
• Agent Skills: Implements the open Agent Skills standard for modular, portable instruction packages. CaveAgent extends the standard with runtime injection (injection.py).
• Multi-Agent Coordination: Control sub-agents programmatically through runtime injection and retrieval. Shared runtimes enable instant state synchronization.
• Streaming & Async: Real-time event streaming and full async/await support for optimal performance
• Execution Control: Configurable step limits and error handling to prevent infinite loops
• Flexible LLM Support: Works with any LLM provider via OpenAI-compatible APIs or LiteLLM
• Type Injection: Expose class schemas for type-aware LLM code generation

Awesome Blogs

We thank these community to post our work.

• CaveAgent让LLM学会了“跑代码”，你能把Agent变成Jupyter里的“老司机”
• Token消耗减半性能满分！状态化运行时管理能力让智能体性能飞升
• Stateful environment to LLMs
• TEKTA-AI

Configuration

Parameter	Type	Default	Description
model	Model	required	LLM model instance (OpenAIServerModel or LiteLLMModel)
runtime	PythonRuntime	None	Python runtime with variables, functions, and types
max_steps	int	5	Maximum execution steps per run
max_history	int	10	Maximum conversation history length
max_exec_output	int	5000	Max characters in execution output
skills	List[Skill]	None	List of skill objects to load
skills_dir	str | Path	None	Directory to discover and load skills from
instructions	str	default	User instructions defining agent role and behavior
system_instructions	str	default	System-level execution rules and examples
system_prompt_template	str	default	Custom system prompt template
python_block_identifier	str	python	Code block language identifier
messages	List[Message]	None	Initial message history
log_level	LogLevel	DEBUG	Logging verbosity level

LLM Provider Support

CaveAgent supports multiple LLM providers:

OpenAI-Compatible Models

from cave_agent.models import OpenAIServerModel

model = OpenAIServerModel(
    model_id="gpt-4",
    api_key="your-api-key",
    base_url="https://api.openai.com/v1"  # or your custom endpoint
)

LiteLLM Models (Recommended)

LiteLLM provides unified access to hundreds of LLM providers:

from cave_agent.models import LiteLLMModel

# OpenAI
model = LiteLLMModel(
    model_id="gpt-4",
    api_key="your-api-key",
    custom_llm_provider='openai'
)

# Anthropic Claude
model = LiteLLMModel(
    model_id="claude-3-sonnet-20240229",
    api_key="your-api-key",
    custom_llm_provider='anthropic' 
)

# Google Gemini
model = LiteLLMModel(
    model_id="gemini/gemini-pro",
    api_key="your-api-key"
)

Contributing

Contributions are welcome! Please feel free to submit a PR. For more details, see CONTRIBUTING.md.

Citation

If you use CaveAgent in your research, please cite:

@article{ran2026caveagent,
  title={CaveAgent: Transforming LLMs into Stateful Runtime Operators},
  author={Ran, Maohao and Wan, Zhenglin and Lin, Cooper and Zhang, Yanting and others},
  journal={arXiv preprint arXiv:2601.01569},
  year={2026}
}

License

MIT License