Code agents

AI MCN 自动推广工具 — 多Agent内容营销自动化，支持掘金/知乎/Dev.to

Deterministic triage of an AI agent's changes — see what it actually touched before you merge.

An honest pre-deployment reliability and security gate for AI agents.

Local-first safety and reliability evaluation framework for AI coding agents.

Independent, auditable verification for coding-agent claims

A secure, monetized Python sandbox for data optimization and Matplotlib rendering via x402.

A green build is not enough: Agentic HIL lets an AI agent probe, flash, reset and exchange UART and CAN traffic with a real STM32 or other embedded target, through policy-gated MCP tools instead of a raw debugger shell, so hardware-in-the-loop tests run unattended in CI.

Stateful LangGraph workflow that reviews Git diffs with a reviewer-verifier loop and deterministic checks.

A status light and a status line for Claude Code, in a cyberpunk terminal panel.

CLI that converts any repo into an agent-ready workspace: detects your stack, scaffolds progressive context and session handoffs, installs a real Claude Code adapter. Extensible plugin architecture for memory backends and coding-agent adapters.

Build reliable coding agents for any OpenAI-compatible /chat/completions endpoint—with spec-driven tools, observability, cache verification, context compaction, and sandboxing. Less glue code, fewer crossed fingers.

A powerful observability and debugging SDK for AI Agents

A local, offline digest of your coding-agent sessions. Zero dependencies.

Harness-agnostic, self-verifying autonomy wrapper for coding agents

Agent work attestation and merge governance: SHA-bound evidence rails for agent-driven development

A proactive memory layer for long-horizon coding agents, it knows when to remind.

Local daemon that bridges Claude Code hooks to AgentMem.

Monitor quota, tokens, and gateway balances for AI coding CLIs.

Parses agent harness session logs into a unified event schema. Python wrapper around the agentminutes Go CLI.

Package for running Agentomics - an autonomous agentic system for ML model development

[Prototype] Filesystem-first durable agent framework built on pydantic-ai v2 — exploring ideas inspired by Eve.dev, fully open source with no vendor lock-in

A vendor-neutral policy-as-code toolkit for governing AI agents and agentic workflows.

A clean Python and CLI interface for authoring and querying SWI-Prolog, built for agents.

CLI security scanner: probes AI-agent HTTP endpoints for secret/prompt leakage, with CI-gating exit codes.

Analyze local Codex, Claude Code, OpenCode, and pi sessions

Tool-agnostic role-to-model mapping for agentic coding workflows

Show what an AI coding agent actually loads from your AGENTS.md files - including the parts silently truncated away.

Run a backlog to completion UNATTENDED — a portable Agent Skill + stdlib harness giving a coding agent durable per-ticket state, an ASK/PARK/HALT autonomy contract, deterministic test-gates, git-backed reversibility and a run report.

Local multi-agent knowledge capture toolkit with CLI and MCP server.

Locally running chat agent with access to Claude Code, Codex, and bash

Let a Python package carry its own agent integrations and install them into whatever agent harness is present.

Audit and scaffold AI harness configs for Claude Code and Hermes Agent — score your setup out of 100

Invokes agent harnesses (Antigravity CLI, Claude Code, Codex CLI) headlessly. Python wrapper around the agentsummons Go CLI.

Qualify repeated categorical AI-agent evidence before it becomes a regression reference

Local containment runtime and firewall for AI coding agents

A local OpenAI-compatible proxy that reduces agent context waste and proves behavior held.

Predict when production AI agents degrade — before they visibly fail. Python SDK.

Multi-agent message broker for AI agent coordination via MCP

独立于具体 Agent 的持久化记忆层 (MCP Server),可被 Claude Code / Qoder / Cursor 复用

Production-grade AI agent orchestration system — Boss, Manager, Coder, Critics, and Worker pipeline

Code-first agent memory SDK for AideMemo (aidememo)

Python implementation of Claude Code - an agentic coding assistant in your terminal

Open AI API compatiable LLM server for Amlogic Device

Keep code adherent to a described architecture; teach coding agents to reason about it.

Ares — AI-native host security investigator for Linux: continuous eBPF telemetry with minute-by-minute AI investigation.

Argus RedTeam local security scan agent — code never leaves your infrastructure

Argus RedTeam local security scan agent — code never leaves your infrastructure

Astra — a terminal AI assistant that understands natural language and drives your dev tools.

Checker and validator for the Barrage language — makes code readable by anyone

Local Apple Silicon LLM Coding Agent Runtime & BaseRT Task Orchestrator with HDAR Receipts

Mobile-first PWA remote control for Claude Code sessions on your Mac

Bharat-Tiny-LLM - Offline Hindi/Hinglish AI assistant for Apple Silicon. `bharat chat` and go. 33.8% token compression via Brahmi injection.

Cloud-native LLM Agent Development and Testing Toolkit

The Cardinal agent-telemetry contract: OTLP emission, initiative resolution, spend-limits delivery, device-code consent.

File-backed mail, MCP tools, push delivery, and authenticated browser messaging for Codex agents

Local analytics for Claude Code project exports

Python SDK for capturing and shipping AI agent traces to a Chronicle server

Claude Code-native failure correction and recovery CLI

Discovers Claude Code capabilities — skills, agents, plugins, MCP servers, hooks — into typed catalogs and an environment report.

A Claude Code token-usage meter for the Windows taskbar

Unofficial tool for Claude Code sessions on disk: make a session visible under your other Claude account on the same machine, or move one between project folders. Command line or GUI. Journaled, reversible, fails closed.

Drive Claude Code from Telegram: live session cards, policy-gated auto-approval, remote approve/deny, mid-run questions, resume, and a Mini App dashboard.

Local queue and auto-continue runner for Claude Code sessions and Codex App tasks.

Drive any wired VIA-compatible RGB keyboard from Claude Code status

Local web dashboard for your installed Claude Code configuration — plugins, agents, skills, commands, hooks, MCP servers, and rules, with usage analytics.

Claude Code rate-limit monitor: floating dongle + dashboard (burn rate, overflow forecast, per-project usage).

Fleet monitoring for Claude Code sessions - TUI dashboard, MCP server, and terminal focus

Local knowledge graph for Claude Code — call graphs, impact radius, and code search via MCP, powered by Tree-sitter and SQLite. Supports Python, JS/TS, Go, Rust, Java, C#, Ruby, C/C++, Swift and more.

Sync Claude Code sessions between two machines over SSH — no cloud.

A fake Claude Code session that looks incredibly productive but does absolutely nothing.

Queue research questions all day. Wake up to answers. Puts your idle Claude Code quota to work while you sleep.

Generator kompletnego, żywego środowiska Claude Code do planowania projektów IT dla software house'u — wizja, ADR, roadmapa, agenci, skille, dowolny stack.

Your Claude Code session sees the 5-hour limit coming, checkpoints its work, and resumes itself when the window resets.

Python backend parity for claude-pty-web-harness: drive Claude Code in a pupptyeer pty and stream its JSONL transcript over the same HTTP/WS protocol as the TS server.

Browser-based code review tool for Claude Code

Blazing-fast local full-text search (SQLite FTS5) for your Claude Code sessions

A doctor for your Claude Code agent skills: find the SKILL.md files Claude silently can't see (over the discovery budget), plus naming, routing and hygiene problems — deterministically, across every installed skill.

A safe, human-in-control companion that waits for Claude Code usage resets, resumes your session, and optionally automates repetitive permission prompts for the active task.

Export Claude Code token usage (per agent/model) to Prometheus, OTLP, or stdout -- parsed from transcripts, works on Max/Pro subscriptions.

Local Claude Code token activity graphs with explicit source accuracy

Cap-aware tmux scheduler & auto-resume for Claude Code — schedule prompts and keep sessions running through the 5-hour usage limit.

Research-agent augmentation layer for Claude Code with SQLite memory, verification MCPs, a proof trunk with NL workflow + Lean reinsurance, and a Textual cockpit TUI.

Reads a startup's codebase and generates a stage assessment, action plan, and hiring/networking recommendations.

AWS CLI that speaks your language. Natural language to AWS operations via Amazon Bedrock.

An MCP server that reviews your code the way a kernel maintainer would: bluntly, technically, without diplomacy.

An autonomous MCP agent that provides token-efficient code analysis, documentation retrieval, and implementation guidance for large library repositories using a dedicated OpenAI-compatible LLM.

Symbol-aware, staleness-tracking RAG over a codebase, exposed over MCP; layers on rag-timetravel.

Auto-generated, budget-bounded codebase maps inside AGENTS.md — spec + stdlib-only reference generator with git-hook, CI, watch, and MCP integrations

AI pair programming in your terminal — modular core (snippet tools, bounded agent loop, permissions)

An unofficial external role runner for the OpenAI Codex CLI

Policy and hook adapters for routing Codex subagents

VCR / time-travel replay for Claude Code and Codex — deterministic offline agent testing

Lightweight agent that backs up network device configs (MikroTik, Ubiquiti) to ConfigStash

Agent loop, harness, and tool dispatch for cortex

Get a second opinion from another AI coding agent. Ask Claude, Codex, OpenCode, CommandCode, or Gemini — from inside the agent you already use.

Fail-closed Custodian policy guard for Claude Code, with unstrippable managed-hook enforcement

AI-security framework — build agents that are safe by construction (placeholder, under active development)

A terminal-first Python coding agent with a bounded model-tool harness.

Claude Code Personal Analytics — a private, local, read-only analytics dashboard for your Claude Code sessions: costs, tokens, tools, skills, coding patterns, and wasted effort.

Sync Claude Code configuration across machines (ccs): collect, apply, and bootstrap from a GitHub payload repo