# agent-base
A lightweight **Agent Runtime Kernel** for building AI agents in Rust.
`agent-base` provides the minimal orchestration layer needed to build custom AI agents — LLM integration, tool dispatch, multi-turn conversation, approval flows, event streaming, and error recovery — all with zero business assumptions.
## Design Principles
- **Clear semantics** — `RunOutcome` explicitly distinguishes `Completed` from `Failed`; events capture the process, the return value captures the final result.
- **Simple state model** — Runtime memory is the source of truth for live sessions; `SessionStore` is an optional persistence adapter.
- **Conservative by default** — On tool failure, the runtime stops by default (`StopOnError`) rather than guessing how to recover.
- **Strategy injection** — All variable behaviors are injected via traits (`ToolErrorRecovery`, `ToolPolicy`, `ApprovalHandler`, `Middleware`), not hardcoded.
## Features
- **LLM Abstraction** — `LlmClient` trait with built-in OpenAI and Anthropic implementations
- **Tool System** — `Tool` trait + `ToolRegistry` for registration and dispatch
- **Approval Flow** — `ApprovalHandler` trait with `AllowOnce` / `AllowAlways` / `Deny` decisions
- **Error Recovery** — `ToolErrorRecovery` trait; defaults to `StopOnError`, opt-in `RetryOnError`
- **Event Streaming** — Structured `AgentEvent` stream for UI, logging, auditing, and debugging
- **Multi-turn Sessions** — `AgentSession` manages message history; `SessionStore` for optional persistence
- **Sub-Agents** — `SubAgentTool` with `Ephemeral` (default) or `Persistent` session policies
- **Context Management** — configurable `ContextWindowManager` for token budget control
- **Middleware** — hooks at `on_user_message`, `on_pre_llm`, and `on_post_llm` for extensions
- **Checkpoints** — structured `Checkpoint` events enable future replay, debugging, and resume
- **MCP Support** — built-in `McpClient` for the Model Context Protocol
- **Skills** — composable capability units with auto-registered tools and on-demand detailed prompts
## Quick Start
### 1. Define a Tool
Any capability you want your agent to have is expressed as a `Tool`:
```rust
use agent_base::{Tool, ToolContext, ToolOutput, ToolControlFlow, AgentResult};
use async_trait::async_trait;
use serde_json::{json, Value};
struct WeatherTool;
#[async_trait]
impl Tool for WeatherTool {
fn name(&self) -> &'static str { "get_weather" }
fn definition(&self) -> Value {
json!({
"type": "function",
"function": {
"name": "get_weather",
"description": "Get current weather for a city",
"parameters": {
"type": "object",
"properties": {
"city": { "type": "string", "description": "City name" }
},
"required": ["city"]
}
}
})
}
async fn call(&self, args: &Value, _ctx: &ToolContext) -> AgentResult<ToolOutput> {
let city = args["city"].as_str().unwrap_or("unknown");
Ok(ToolOutput {
summary: format!("Weather in {}: 22°C, sunny", city),
raw: None,
control_flow: ToolControlFlow::Continue,
truncated: false,
})
}
}
```
### 2. Build the Agent
```rust
use std::sync::Arc;
use agent_base::{
AgentBuilder, AgentEvent, AgentResult, RunOutcome,
OpenAiClient, StopOnError,
};
#[tokio::main]
async fn main() -> AgentResult<()> {
let llm = Arc::new(OpenAiClient::new(
std::env::var("OPENAI_API_KEY").unwrap(),
"gpt-4o".into(),
None,
));
let mut runtime = AgentBuilder::new(llm)
.system_prompt("You are a helpful weather assistant.")
.register_tool(WeatherTool)
.build();
let session_id = runtime.create_session();
let (events, outcome) = runtime.run_turn_stream(
session_id,
"What's the weather in Tokyo?",
).await?;
for event in &events {
match event {
AgentEvent::TextDelta { text, .. } => print!("{}", text),
AgentEvent::ToolCallStarted { tool_name, .. } => {
println!("\n[Calling tool: {}]", tool_name);
}
AgentEvent::ToolCallFinished { summary, .. } => {
println!("[Tool result: {}]", summary);
}
AgentEvent::RunFinished { .. } => println!("\n[Done]"),
_ => {}
}
}
assert_eq!(outcome, RunOutcome::Completed);
Ok(())
}
```
### 3. Handle Tool Errors
By default, tool failures stop the run. For self-healing agents (e.g. code agents that retry compilation), inject `RetryOnError`:
```rust
use agent_base::RetryOnError;
let mut runtime = AgentBuilder::new(llm)
.register_tool(MyTool)
.error_recovery(Arc::new(RetryOnError)) // ← retry on failure
.build();
```
### 4. Add Approval for Sensitive Tools
```rust
use agent_base::{
ApprovalHandler, ApprovalRequest, ApprovalDecision,
ToolPolicy, RiskLevel,
};
struct MyApprovalHandler;
#[async_trait::async_trait]
impl ApprovalHandler for MyApprovalHandler {
async fn approve(&self, _req: ApprovalRequest) -> AgentResult<ApprovalDecision> {
// Ask user via UI, CLI, etc.
Ok(ApprovalDecision::AllowOnce)
}
}
struct MyToolPolicy;
impl ToolPolicy for MyToolPolicy {
fn evaluate_approval(&self, tool_name: &str, _args: &Value, _json: &str)
-> Option<ApprovalRequest>
{
if tool_name == "dangerous_tool" {
Some(ApprovalRequest {
title: "Confirm action".into(),
message: format!("Execute `{}`?", tool_name),
risk_level: RiskLevel::Sensitive,
..Default::default()
})
} else {
None // auto-allow
}
}
}
let mut runtime = AgentBuilder::new(llm)
.register_tool(DangerousTool)
.tool_policy(Arc::new(MyToolPolicy))
.approval_handler(Arc::new(MyApprovalHandler))
.build();
```
### 5. Use a Sub-Agent
```rust
use agent_base::SubAgentTool;
// Build a sub-agent runtime
let sub_llm = Arc::new(OpenAiClient::new(key, model, None));
let sub_runtime = AgentBuilder::new(sub_llm)
.system_prompt("You are a math expert.")
.build();
// Wrap it as a tool
let math_tool = SubAgentTool::new(
"calculate",
"Delegate math problems to a math expert",
sub_runtime,
);
// Register in the parent agent
let mut parent = AgentBuilder::new(parent_llm)
.register_tool(math_tool)
.build();
```
Each sub-agent call creates a fresh session by default. Use `SubAgentTool::with_persistent()` to share context across calls.
## Examples
```bash
# Configure API key
cp .env.example .env
# Edit .env with your OPENAI_API_KEY or ANTHROPIC_API_KEY
# Run the REPL example
cargo run --example repl
# Run the SubAgent demo
cargo run --example subagent_demo
# Run the MCP demo
cargo run --example mcp_demo
# Run the Skill demo
cargo run --example skill_demo
```
## What agent-base Does NOT Do
- ❌ No SSH, filesystem, or database tools
- ❌ No workflow DAG / multi-agent orchestration engine
- ❌ No memory / RAG framework
- ❌ No terminal UI or approval dialogs
- ❌ No heavy persistence or transaction system
Business-specific tools and strategies belong in **upper layers** (`ops-agent`, `db-agent`, etc.).
## Typical Layering
```
ops-agent / db-agent / browser-agent ← Business agents
└── agent-base ← Lightweight Runtime Kernel
```
## v1 Semantics
| `run_turn_*` → `AgentResult<RunOutcome>` | `Ok(Completed)` = success, `Ok(Failed)` = finished with error |
| `AgentEvent::RunFinished` | Process ended — final status is in `RunOutcome` |
| Tool failure → defaults to `StopOnError` | Inject `RetryOnError` for self-healing agents |
| SubAgent → defaults to `Ephemeral` | Use `with_persistent()` for shared context |
| Session → memory is source of truth | `SessionStore` is an optional persistence adapter |
## Stability
This project is in early development (v0.1.0). The core abstractions are settling but not yet frozen. Expect minor API changes as the ecosystem evolves.
## License
MIT