Code agents

AI-powered Python debugger — run a broken script, get root cause, explanation, and fix

Spot AI-hallucinated and slopsquatting Python dependencies, typosquats, and hardcoded secrets.

An open-source AI agent that automatically investigates DevOps incidents and identifies root causes.

Run large LLMs across two NVIDIA DGX Sparks with vLLM — model picker, one-command start/stop, and an Anthropic-compatible endpoint for Claude Code.

Installer and CLI for the local DiffusionGemma coding agent runtime

Agents de code spécialisés dans le terminal : un modèle LLM par agent, sur n'importe quel back-end.

The dotagents CLI: install and manage a portable ~/.agents config for AI coding agents

An MCP that add terminal capabilities to an agent

Production-grade DSPy library for explainable HS code classification.

Evaluation-Driven Development (EDD) tools and helper libraries for AI Agent development

LLM evaluation framework for developers

A practical PRAGMA model implementation for tokenized tabular event histories

FlowJet (fj) — one-shot CLI coding agent powered by soothe-nano

Run Google ADK (Agent Development Kit) agents on Flyte.

Predictive cost, policy, and runaway-session guardrails for AI coding agents

A modern multi-repository GitHub TUI with Claude Code integration

A spec-driven development PM agent that orchestrates multiple coding CLIs using GitHub Issues as state

Code Property Graphs — Joern/Semgrep extraction, AI vulnerability enrichment, risk scoring

Scaffold a self-improving agent governance framework (meta-skills, hooks, memory hierarchy) for Claude Code.

Reusable helpers for HEP large-language-model generated analysis code.

Local-dev MCP sidecar for exposing selected Hermes Agent capabilities.

Agent development kit for building self-tuning AI agents with HippoDb biological memory

HiveLLM binary RPC (Thunder) - wire v1 codec, family profiles, sync/async multiplexed clients

Local HWP/HWPX retrieval for Codex and Claude Code over MCP

Production-grade, non-breaking live debugger and telemetry agent for Python.

Where fantasy meets functionality. The best of many worlds, refined into one intelligent agent.

The code for the book 'An Illustrated Guide to AI Agents'

inis.run CodeAct provider for Microsoft Agent Framework

jrag (formerly java-codebase-rag) — MCP server + jrag CLI for semantic + structural search over Java/Kotlin codebases

Decompose a task with Codex and dispatch the subtasks to Jules.

Give your LangGraph agent a June memory — mirror chat history from any LangGraph checkpointer into a searchable, cited June knowledge canvas. Three commands, no code changes.

Project-centered Pi coding agent and persistent Chatrooms for JupyterLab

Collaborate with Claude on Jupyter notebook code and documentation

KRISH-AGENT INFINITY: The ultimate AI coding assistant. 1 billion% better with GODMODE, HYPERINFINITY, and QUANTUM UNIVERSE capabilities.

Baponi sandbox provider for LangChain Deep Agents

Drop-in ChatAnthropic replacement backed by the Claude Code CLI — use your Claude Pro/Max subscription, no API key needed

Design-time low-code flow builder for the agentplane platform: edit, validate, publish FlowDefinitions.

Self-healing runner for Python projects: catches crashes, remembers them, and patches them on the fly with a headless Claude agent.

Batch-API-native coding agent: plan realtime, execute on 24h-SLA batch APIs, optimize like a query engine.

A dev tool for manually driving llm-agents-from-scratch's SupervisedTaskHandler one call at a time, over HTTP, via a React frontend.

LaC (LLM as Code) - reference engine. Declarative behavior for LLM apps: law in files, perimeter in code.

Cross-vendor LLM quota / budget self-check CLI. Stable 0/1/2/3 exit-code contract lets any AI agent gate its work before burning through budget. Supports MiniMax Token Plan, OpenRouter credits, self-hosted LiteLLM proxy spend, and self-logged JSONL (any LLM vendor).

OpenRouter LLM benchmark for repair-agent and validator-agent model selection.

Flight recorder for LLM agents: trace every call, replay runs deterministically, diff runs to find where behavior diverged.

Local LLM token efficiency middleware — zero cloud, zero tracking, coaches developers on token waste in the terminal

LLM plugin for Z.AI / GLM models with Coding Plan support

LLM plugin for Z.AI / GLM models with Coding Plan support

Interactive LLM-driven automated algorithm design with evolutionary optimization

A lightweight, personal Claude Code-style agentic CLI for LOCAL LLMs (LM Studio).

Scored tooling katas, habit rules, and pre-baked warm-start sessions for LLM coding agents

LLM Surgery

Shared AWS Bedrock internals (SigV4 signing, event-stream decoding, Anthropic pricing) for lmux

Use the OpenAI Python SDK with your own ChatGPT subscription

Launch each project's AI coding agent in its own terminal, auto-tiled across every monitor

Observability for AI agent loops: trace trees, metrics, loop detection, cost tracking. Zero dependencies.

Deterministic microscopy image analysis for AI agents, over MCP. Point your Claude/Cursor/Codex agent at an image and it segments, counts, and measures cells — Fiji/ImageJ-equivalent, every number traceable to a tool, never guessed.

A developer-facing memory layer for AI agents (MCP server).

A minimal coding agent

General purpose agent harness inspired by Claude Code

A compact transactional runtime for verified coding agents.

A lean, snappy fullscreen terminal client for Claude Code

Closed-loop, self-improving agent engine: Claude Code, Google Antigravity and OpenAI Codex compete for the Editor seat via blind-tournament judging, a deterministic referee (Haskell decision kernel + verified Python fallback) polices the flow, and git + an objective fitness signal arbitrate every change. Persistent sessions, context-rotation memory, AST code graph, temporal experiment knowledge graph.

Local-first multilingual RAG for code and technical docs. In-process hybrid retrieval (vector + full-text) with FTS for Cyrillic and other non-ASCII scripts.

Themed, non-repeating nicknames for Claude Code subagents — a userspace port of Codex's nickname_candidates

All-in-one RAG framework for Nepali Documents.

Ness Agent is an experimental, hackable agent harness.

NextPy AI Agent - Intelligent coding assistant for NextPy framework

Archive AI coding-agent sessions (Claude Code, Codex, Gemini, Cline, OpenCode, Hermes) as index-safe Markdown notes in an Obsidian vault.

Repository-owned knowledge maintenance for coding agents

Meta-package: installs the full mitodl agent-kit toolkit (agent-config-kit's agent-kit CLI + witan + witan-code)

A local multi-agent code reviewer

Persistent, branch-aware, git-anchored memory & context layer for coding agents.

A pure-stdlib hybrid Code-as-Action scientific research agent.

Install agent tool packages into global or project OpenCode configurations

Agentic system optimization: coding agents reason over parameter meaning and trial history.

Claude Code, overclocked — multi-account hot-swap, context-threshold handoff, instrumented statusline, and model routing for multi-agent workflows

Python SDK for the Persona.ai Developer Platform API. Server-side only — never bundle into client-side/browser code.

Deterministic orchestration of coding-agent pipelines (supervised subprocesses, DAG, bounded re-runs).

Python port of @earendil-works/pi-coding-agent — Coding agent SDK (core only, no TUI)

AI Agent that auto-generates Postman Collections from backend code

Local RAG MCP server for code and documentation

Catch AI coding agents cheating on their work — verdict on facts, not opinions.

The official Claude Code agent — same as Anthropic's, full coding toolset. GPU via providers.

The official OpenAI Codex agent — same as OpenAI's, full coding toolset. GPU via providers.

General coding assistant — one harness across Claude / ChatGPT / Gemini. No PWM data.

Jason-style AgentSpeak for Python. Fork of niklasf/python-agentspeak, a Python interpreter for the Jason (jason-lang.github.io) agent-oriented programming language.

Local MCP server exposing QpiAI Quantum SDK tools to agent harnesses

A debugger for RAG applications: diagnoses retrieval, context, and grounding failures.

Lightweight JSONL tracing for debugging failed RAG answers.

RAG indirect prompt injection guard: scan retrieved documents for prompt-injection (instruction override, system/tool-prompt manipulation, exfiltration, invisible-Unicode/homoglyph smuggling) in English and Turkish, and quarantine poisoned docs before they reach the model.

AI Coding Agent

A simple, extensible toolkit for building advanced multi-agent Retrieval-Augmented Generation (RAG) pipelines in a few lines of code.

RAGL — a memory-safe, flexible-syntax, AI-native programming language

Self-hosted RAG engine: hybrid dense+sparse retrieval, 28-format ingestion (docs, URLs, images, audio, video), 6 vector backends, 8 embedding providers, 12+ generation providers with automatic fallback, query rewriting (HyDE/contextual/multi-query), structured JSON output, real per-call cost tracking, and standalone retrieval via retrieve(). Composable with the optional ragleap-graph package for knowledge-graph-augmented retrieval. BYOK, no vendor lock-in, no hardcoded model defaults.

Rein — keep your Claude Code agents lean and in control. Token-lean, stack-aware agentic dev flow.

AI-Powered Codebase Bug & Vulnerability Scanning CLI

Local-first AI code agent in your terminal. Inspired by Claude Code and Codex CLI.

Record-replay debugger for AI agent runs

Robot Framework libraries for testing the agentic stack: MCP servers, Agent Skills, SubAgents, and Hooks - deterministically, with an LLM judge, or with a coding agent.

A small example package