ReloadedCode

Production-grade coding agent tools in Rust. ~10 MiB. No TUI. Embed it anywhere.

Get Started · Documentation · API Reference · Examples

---

Why this project?

ReloadedCode started as "an OpenCode for servers." Headless, sandboxed, and cheap to host for non-commercial use.

OpenCode is a great interactive coding agent, but it's a ~305 MiB TypeScript application that runs as a separate process. What if you need those same tools for a server? A Discord bot? A CI pipeline? A custom product?

ReloadedCode ships the same agent tools as a Rust library. Shell sandboxing. Default-deny permissions. ~10 MiB footprint.

	OpenCode	ReloadedCode
Language	TypeScript	Rust
Memory	~305 MiB	~13 MiB
Interface	TUI / Desktop / IDE	Library (headless)
Agent format	Markdown + YAML	Similar format
Embeddable	HTTP API	Rust crate

Features

• 10 Built-in tools - read, write, edit, glob, grep, bash, webfetch, todoread, todowrite, task
• Agents similar to OpenCode - load agent markdown files with YAML frontmatter
• Multi-agent delegation - orchestrator pattern with depth-limited task chains
• Linux sandboxing - bubblewrap profiles for shell isolation (Public Bot + Trusted Maintenance)
• Path security - restrict file access with allowed directories and glob-based rules
• Model catalog - sync models.dev with ETag caching (~3000 models, ~24 KiB cache)
• Permissions - default-deny with last-match-wins rules and wildcard patterns
• Optimized system prompt - auto-generated, ~2000 tokens full / ~560 search-only
• Async + Sync - every tool compiles as tokio async or blocking. Zero overhead.
• Framework-agnostic core - use the serdesAI integration or bring your own framework
• 15 LLM providers - OpenAI, Anthropic, Google, Groq, Mistral, Ollama, Azure, Bedrock, and more
• Semver-guaranteed API - 6-platform CI matrix, 11 semver surfaces, clippy -D warnings

Quick Start

[dependencies]
reloaded-code-serdesai = "0.2"

1. Create an agent file (markdown + YAML frontmatter similar to OpenCode):

---
name: coder
mode: all
description: A coding agent that can read, search, and edit files.
permission:
  read: allow
  write: allow
  edit: allow
  glob: allow
  grep: allow
  bash: allow
  webfetch: allow
  task: deny
---

You are a coding assistant. Use the available tools to complete the user's task.

2. Load the catalog, build the agent, and run:

use reloaded_code_agents::{AgentCatalog, AgentLoader, AgentRuntimeBuilder};
use reloaded_code_core::CredentialResolver;
use reloaded_code_models_dev::ModelsDevCatalog;
use reloaded_code_serdesai::{AgentBuildContext, AgentDefaults};
use std::{path::PathBuf, sync::Arc};

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    // Load agents from the "agents" directory.
    let mut catalog = AgentCatalog::new();
    AgentLoader::new().add_directory(&mut catalog, "./agents")?;

    // Supports any model from https://models.dev
    let load_result = ModelsDevCatalog::load().await?;

    let runtime = AgentRuntimeBuilder::new()
        .catalog(catalog) // Default model if not specified by agent.
        .defaults(AgentDefaults::with_model("synthetic/hf:MiniMaxAI/MiniMax-M2.5"))
        .build()?;

    let build_context = AgentBuildContext::new(
        Arc::new(runtime),
        Arc::new(load_result.catalog),
        Arc::new(CredentialResolver::new()),
    );

    let agent = build_context.build("coder")?;
    let response = agent.run("Find all TODO comments in src/", ()).await?;
    println!("{}", response.output());
    Ok(())
}

Crate Map

Crate	Version	Description
reloaded-code-core	0.2	Framework-agnostic tool implementations, path resolvers, permissions, system prompt builder
reloaded-code-agents	0.1	agent markdown loader similar to OpenCode, typed catalog, runtime builder
reloaded-code-serdesai	0.2	SerdesAI framework integration, tool adapters, 15 provider bridges, task delegation
reloaded-code-bubblewrap	0.1	Linux bubblewrap sandbox profiles (Public Bot + Trusted Maintenance)
reloaded-code-models-dev	0.1	models.dev catalog sync with ETag caching and offline fallback

Examples

# Basic agent setup with all tools
cargo run --example serdesai-basic -p reloaded-code-serdesai

# Sandboxed file access (restricted to allowed directories)
cargo run --example serdesai-sandboxed -p reloaded-code-serdesai

# Sandboxed bash execution (Linux, requires bubblewrap)
cargo run --example serdesai-sandboxed-bash --features linux-bubblewrap -p reloaded-code-serdesai

# Agent catalog loading from markdown files
cargo run --example serdesai-agents -p reloaded-code-serdesai

# Multi-agent task delegation (orchestrator + reader sub-agent)
cargo run --example serdesai-task -p reloaded-code-serdesai

Documentation

• Documentation Site - guides, architecture, examples
• reloaded-code-core README
• reloaded-code-agents README
• reloaded-code-serdesai README
• reloaded-code-bubblewrap README
• reloaded-code-models-dev README
• Sandbox profiles and operator checklist
• API Reference (docs.rs)

Contributing

Contributions are welcome! Please ensure all tests pass and the code follows our guidelines.

License

Licensed under Apache 2.0.