llm-coding-tools-core

Lightweight, high-performance core types and utilities for coding tools - framework agnostic.

Overview

This crate provides the foundational building blocks for coding tool implementations:

• ToolError - Unified error type for all tool operations
• ToolResult<T> - Result type alias using ToolError
• ToolOutput - Wrapper for tool responses with truncation metadata
• Utility functions for text processing and formatting
• context module - LLM guidance strings for tool usage

Features

• tokio (default): Async mode with tokio runtime. Enables async function signatures.
• blocking: Sync/blocking mode. Mutually exclusive with tokio/async.
• async: Base async signatures (internal). Requires a runtime; use tokio instead.

The async and blocking features are mutually exclusive - enabling both causes a compile error.

Future runtimes (smol, async-std) can be added following the same pattern as tokio.

Usage

use llm_coding_tools_core::{ToolError, ToolResult, ToolOutput};
use llm_coding_tools_core::util::{truncate_text, format_numbered_line};

Context Module

The context module provides embedded strings containing usage guidance for LLM agents. These can be appended to tool descriptions or system prompts.

Path-based tools have two variants:

• *_ABSOLUTE: For unrestricted filesystem access (absolute paths required)
• *_ALLOWED: For sandboxed access (paths relative to allowed directories)

use llm_coding_tools_core::context::{BASH, READ_ABSOLUTE, READ_ALLOWED};

// Non-path tools have a single variant
println!("{}", BASH);

// Path-based tools have absolute and allowed variants
println!("{}", READ_ABSOLUTE);
println!("{}", READ_ALLOWED);

Available context strings:

• BASH, TASK, TODO_READ, TODO_WRITE, WEBFETCH - standalone tools
• READ_ABSOLUTE, READ_ALLOWED - file reading
• WRITE_ABSOLUTE, WRITE_ALLOWED - file writing
• EDIT_ABSOLUTE, EDIT_ALLOWED - file editing
• GLOB_ABSOLUTE, GLOB_ALLOWED - pattern matching
• GREP_ABSOLUTE, GREP_ALLOWED - content search

Design Principles

• No framework-specific dependencies, plug and play into any LLM framework/library
  • See llm-coding-tools-rig for an integration example with rig
• Minimal dependency footprint
• Performance-oriented (optimized) with zero-cost abstractions