human-in-the-loop — independent software & tools
-
aipermission
Local-first permission gateway for AI agents: connector-based SSH, Postgres, and Redis access with scoped tokens, human approval, audit logs, and encrypted local credentials.
-
huqan
Local-first, deterministic verification layer for AI claims, memory writes, and risky actions.
-
sales-agent-showcase
Clean-room showcase of a multi-marketplace resale automation system (sanitized): adapter-pattern cross-posting with retries, idempotency, and tests.
-
AI-Employee-Silver
Automate Gmail and LinkedIn tasks with AI drafts, human approval, and scheduled posting
-
KarvyLoop
The agent that learns how you decide - and proves it by never deciding for you. A local-first, loop-native runtime for a team of AI agents on your machine.
-
druks
Autonomous software delivery on your own Claude/Codex subscriptions — durable runs, human gates, sandboxed agents. Druks is the self-hosted home for durable agent apps.
-
Cheers
A Slack-style hub for humans and AI agents — real-time channels, external agent bots, file-aware chat, and channel memory. Open-source & self-hostable & MacOS & IOS.
-
StratumCode
AI coding agent that shows its work — every step visible, every patch approved by you. Build projects brick by brick, stay in control.
-
ai-flywheel-spec
The normative specification for an evidence-driven AI operating model that learns from execution, governs adaptation, validates improvements, and compounds reusable knowledge over time.
-
io-harness
An embeddable agent runtime for Rust. Any task, any provider, in your process — with a permission boundary, a sandbox, and a durable trace you own.
-
proxima
Self-hosted, single-user control plane for the AI agents you already own. Work hands-on in Chat or delegate to the Alpha orchestration desk, run reviewable jobs and recipes, and review artifacts natively. Runner-agnostic over ACP: Claude Code, Codex, Grok, Hermes, Pi. No bundled model or credentials.
-
ai-hiring-workflow-engine
Analyze resumes against job descriptions using Claude AI to generate match scores, gap insights, and improvement actions for recruiters and job seekers.
-
harness-books
Explore harness engineering books for building bounded, accountable code workflows in terminals, repos, permissions, and team systems
-
Open-Arsenal
Open Arsenal specs can be self-hosted on private or public networks, with free use under AGPL-3.0+ requiring Section 7 attribution. Optional Specification Branding Licenses allow use without attribution, with pricing based on spec, scope, and network size.
-
glance
Agents deploy, you comment, agents fix. A self-hosted human-in-the-loop review loop for coding agents — $0 on Cloudflare's free tier.
-
signal-sweep
Human-gated presence tooling for open-source projects. Module 1: find the GitHub threads your docs already answer.
-
agent-os-langgraph
LangGraph coding-agent orchestrator with HITL, SQLite resume, 3-tier routing, MCP, and Claude/Codex subscription backends.
-
page-agent-sdk
给网页一个会改页面的 AI 助手。一行代码挂载对话框，AI 通过工具按 schema 安全读写页面数据，实现「对话式」搭建/编辑/运维。
-
contenox
Governance for AI agents: one file declares what may run, what needs a human, and what starts work. Enforced by the task engine; the terminal UI and CLI are reference surfaces. Any LLM provider, on your machine.
-
md-redline
Inline review comments for markdown specs. Built-in MCP server hands feedback directly to your AI agent.
-
pitchbox
Self-hosted, human-in-the-loop outreach agent. The agent researches and drafts; you approve before anything is sent.
-
AILIENANT
Ailienant is a structured engineering agentic OS for VS Code. No vibe coding or unauditable black boxes. Built for engineers demanding total control. We provide a live codebase knowledge graph, MCP integration, and total model/budget control. It autonomously plans, executes, and self-corrects with spec-driven rigor and a Zero-Trust ideology.
-
LIA-Assistant
Smart multi-agent conversational assistant with LangGraph orchestration, Human-in-the-Loop, enterprise-grade observability, and full i18n support (6 languages)
-
aria-knowledge
Applied reasoning, persistent agent memory, context engineering, and human-governed trust for AI coding agents — Claude Code, Claude Cowork, Codex, Cursor, Antigravity. Compatible with Obsidian.
-
totem
Rules you can enforce, state you can derive, context you can query: plain files in your repo, under whichever coding agent you use.
-
red-team-blue-team-agent-fabric
AI agent security harness for adversarial testing: 606 executable tests on main, 603 in the v4.15.0 release, across MCP, A2A, x402/L402, decision governance, benchmark integrity, human-in-the-loop, skill supply chain. Commit-pinned OWASP Agentic v1.1 T1-T17 (13 direct, 4 partial), AIUC-1 2026-Q1/Q2 crosswalk 19/20, NIST AI 800-2 aligned.
-
agent-knock-knock
Control local Codex and Claude Code from OpenClaw through shared tmux terminals, with seamless human-agent handoff.
-
synentra
Intent‑aware governance gateway for autonomous AI agents that interact with enterprise APIs.
-
leashd
Safety-first agentic coding framework. Three-layer safety pipeline (sandbox, YAML policies, human-in-the-loop approval) for AI coding agents. Pluggable runtimes (Claude Code, Codex), autonomous task orchestrator, full audit trail.