A Rust framework for building composable AI agent runtimes.

English readme | 中文说明 | 日本語の説明

Anda is a Rust framework for building AI agents that can combine models, tools, memory, and other agents into a single runtime. It focuses on composability, type-safe extension points, asynchronous execution, and practical runtime control.

The core engine lets developers register agents and tools, route model requests by capability labels, call local or remote functions, isolate context state, and add optional persistence or memory layers when an application needs them.

1. Composable agents and tools Agents and tools are registered through stable traits and function definitions, so specialized components can be combined into larger workflows without hard-coding one application shape.

2. Model routing The engine can route completion requests through labeled model tiers such as primary, pro, flash, or lite, while provider-specific adapters stay behind a common request and output contract.

3. Runtime orchestration CompletionRunner handles iterative model turns, tool calls, agent calls, usage accounting, artifacts, steering messages, follow-up messages, and cancellation.

4. Scoped execution context BaseCtx and AgentCtx provide isolated state, cache, object storage, HTTP calls, signed calls, cancellation, and child contexts for each agent or tool.

5. Extensible memory and skills Optional extensions provide conversation storage, KIP-based memory tools, filesystem access, shell execution, fetch, notes, todos, and file-backed skills.

Documents:

• Anda Architecture

anda/
├── anda_cli/              # Command-line interface for Anda engine servers
├── anda_core/             # Core traits, types, and runtime contracts
├── anda_engine/           # Agent runtime, orchestration, contexts, models, and extensions
└── anda_engine_server/    # HTTP server for serving one or more Anda engines

Use anda_cli and anda_engine_server to run and interact with configured engines.

• Build custom agents and tools with the anda_core traits.
• Extend anda_engine with reusable runtime features.
• Improve model adapters, context capabilities, memory integrations, and server APIs.

• Anda Brain: Persistent memory and cognition product built on the Anda framework.
• Anda Bot: Personal AI assistant and application runtime built on the Anda framework.

• KIP: Knowledge Interaction Protocol used by Anda memory tools.

Copyright © 2026 LDC Labs.

ldclabs/anda is licensed under the MIT License. See LICENSE for the full license text.