Code agents

Share code with LLMs via Model Context Protocol or clipboard. Rule-based customization enables easy switching between different tasks (like code review and documentation). Code outlining support is included as a standard feature.

A small CLI tool to collect source files from a repository into a single Markdown document, ready to share with an LLM.

Turn your entire codebase into a single prompt-ready text file.

Scope your AI context before it explodes

Create an LLM XML context document from an llms.txt file

LLM plugin to access Devin API

LLM development debug layer - every API call recorded, nothing lost. Reasoning tokens, prompt diff, session timeline, LangChain/LlamaIndex integration.

LLM plugin to create a hierarchical summary description of a project

Deterministic codebase context retrieval for LLMs

A light-weight Python library 🐍 for deploying DAG chains of customizable Large Language Model operations. Leveraging the asyncio framework, LLM-Director allows developers to create, manage, and orchestrate actions and processes in an efficient and scalable manner. With its unique structure and integration with FastAPI, LLM-Director offes a robust solution for building applications, API endpoints, and workflow automation systems.

AI-powered documentation generator for code projects

Code to generate docstrings for Python code using GPT-4 etc.

Debug plugin for LLM

Use LLM to edit source files

Drop-in observability for LLM applications — quality scoring, hallucination detection, cost tracking, and agent run debugging.

MCP server for agentic LLM evaluation: jury scoring, agent tracing via OpenTelemetry, document-grounded QA generation, PDF reports.

LLM + evolutionary algorithms to optimize programs in multiple languages and domains

LLM explains stack trace for all exceptions. It automatically takes your input code and the error and tries to give a solution.

LLM Expect is a minimalist, developer-first SDK for testing LLM-powered Python functions using structured JSONL datasets.

A prompt graph definition and execution library for LLM app development.

LLM plugin to access Google's Gemini via Code Assist API with OAuth

A lightweight framework that connects LLMs to a virtual computer (Docker-based sandbox) to build general-purpose agents

A collection of commonly used code

Deterministic Unicode-aware input hardening for LLM applications.

Intercept and analyze LLM traffic from AI coding tools

LLM Development Framework - Spec-driven development for AI-assisted engineering

Local web UI for browsing and managing LLM CLI conversation history (currently: Claude Code).

LLM-oriented CLI linter — catch quality drift before it hardens (Rust binary).

Set logging.DEBUG while running LLM

multi-agent development

Personal conversation knowledge base: import, search, and analyze conversations from ChatGPT, Claude, Gemini, and Claude Code

Karpathy-style LLM wiki from your Claude Code, Codex CLI, Cursor, and Obsidian sessions

Access OpenAI models via a Codex subscription

Access OpenAI models via a Codex subscription

Robust search/replace file editing for LLM-driven code changes

A generic JSON prettier and interactive HTML viewer, built especially for parsing embedded JSON from LLM payloads and responses.

Generate LLM-friendly summaries of Python and Go codebases

Secure Python code execution sandbox for LLM-generated code

Unified LLM usage management — API proxy, session diagnostics, multi-CLI orchestration.

Typed LLM wiki graph pipeline for research and development projects

Exponential backoff with full jitter for LLM API calls. Sync + async. Built-in retryable-code presets for Anthropic, OpenAI, Bedrock, Gemini. Zero runtime deps.

A local LLM-powered code review CLI tool

Time-travel debugger for AI agents. Record any production run, replay any failure.

Multi-LLM router MCP server — smart complexity routing, budget-aware model selection, 20+ providers (Claude, OpenAI, Gemini, Ollama, etc.)

LLM plugin exposing the RovoDev model for the llm CLI

Recursive self aggregation

LLM Sandbox is a lightweight and portable sandbox environment designed to run large language model (LLM) generated code in a safe and isolated mode.

CLI + MCP server for detecting, classifying, and redacting embedded LLM agent instructions in documents, source code, and web pages.

Evidence-first, LLM-augmented static code analysis: graph indexing, MCP server, audit workflows

Local debugger for LLM API calls — waterfall breakdown of TTFT, cost, and token speed

Cognitive Security Middleware - The 'Electronic Stability Program' (ESP) for Large Language Models. Bidirectional containment system with defense-in-depth architecture (6 validation layers), stateful tracking, and mathematical safety constraints. Validated against Unicode/encoding attacks, pattern evasion, multilingual/polyglot attacks (12+ languages including Basque, Maltese), and memory/session attacks. Protocol-based hexagonal architecture with LangChain integration.

AI-powered security code reviewer using a fine-tuned Qwen2.5-Coder LLM

Track, visualize, and optimize LLM API spending. Two lines of code, zero config.

A package for generating code using various AI models.

Comprehensive testing suite for LLM evaluation: hallucination detection, consistency, robustness, safety, and multi-language code generation assessment.

Token-efficient code analysis for LLMs. 5-layer stack: AST, Call Graph, CFG, DFG, PDG. 95% token savings. 17 languages.

The foundational shared event schema for the LLM Developer Toolkit

wrapper for cased-kit's code analysis toolset

OpenTelemetry instrumentation for Google ADK (Agent Development Kit).

从 YAML 配置生成标准化的 AI 协作规则文档 (llm.txt)

Local LLM coding usage collector and Feishu Bitable sync

AI-powered linting tool for code quality and validation

Production-grade security sandbox for executing untrusted Python & JavaScript code generated by LLMs using WebAssembly and WASI

Cross-platform LLM Wiki agent initializer and platform adapter.

MCP server + Claude Code skills for Karpathy-style LLM wikis: persistent markdown knowledge bases your agent grows over time.

LLM Woodpecker - AI Model Debugging Tool with Chat and Raw JSON modes

Quick way to access LLM output in CLI and run Code Agents

LLM4Data is a Python library designed to facilitate the application of large language models (LLMs) and artificial intelligence for development data and knowledge discovery.

A Linter that uses LLM to analyze code

K9s-inspired observability control tower for LLM & MCP Servers

Run llm functions with just inline documentation and no code

A framework for simplifying the development of AI systems based on LLMs.

High performance code search for large codebases

A Python library for creating interactive web maps, optimized for LLM-assisted development.

No BS utilities for developing with LLMs

LLMs at your service

A modular LLM application development framework

llmcc brings multi-depth architecture graphs for code understanding and generation.

Llmcode is AI pair programming in your terminal

Open-source AI agent runtime for any LLM — production-grade coding agent with multi-layer memory, multi-agent orchestration, and defense-in-depth security

Installer for the llmcode AI pair programming CLI tool.

A Python library for extracting code snippets using LLMs

LLMCore: Essential tools for LLM development - models, prompts, embeddings, agents, chains, and more.

Active memory and executive control for AI coding agents. Classifies user prompts against a curated tripwire store, injects the matched lessons into the agent's working context via a Claude Code UserPromptSubmit hook, before the agent reasons about the task.

Specwright - Python framework for LLM-assisted development with runtime spec validation

A simple interface to models.dev, the open database of AI model specs, pricing and capabilities

Structured debug snapshots for LLM-assisted debugging

Tools for LLM applications development

git diff for LLM prompts

LLM Development Kit for common APIs

Find LLM cost leaks before your bill does. Static analysis for Anthropic and OpenAI client code.

Security analysis framework for LLM-generated content in software development

llmer is a lightweight Python library designed to streamline the development of applications leveraging large language models (LLMs). It provides high-level APIs and utilities for parallel processing, runtime management, file handling, and prompt generation, reducing the overhead of repetitive tasks.

Reserved name — lldesign: UI/Web/diagram design tooling for the FullSense LLM family. Reference implementation in development.

Reserved name — lleval: LLM eval framework for the FullSense LLM family (on-prem + cloud unified A/B, progressive size matrix, honest disclosure). Reference implementation in development.

Reserved name — lltrade: trading research for the FullSense LLM family (PAPER-TRADING ONLY, v0.x). Reference implementation in development.

Execute LLM-Generated Python Code Automatically

A Python client for interacting with llamafile local LLM instances

A python package for developing AI applications with local LLMs.