by os-factory · MCP Server · ★ 90
Last updated: · Indexed by AgentSkillsHub · Auto-synced every 8h
HAR: open agent harness (CLI + MCP) for coding agents. Isolated worktrees, deterministic verify, software factory workflows for Claude Code, Cursor, and Codex.
| Stars | 90 |
| Forks | 11 |
| Language | TypeScript |
| Category | MCP Server |
| License | Apache-2.0 |
| Quality Score | 52.2156973113354/100 |
| Open Issues | 6 |
| Last Updated | 2026-09-14 |
| Created | 2026-06-28 |
| Platforms | claude-code, cli, codex, mcp, node |
| Est. Tokens | ~13k |
These tools work well together with har for enhanced workflows:
Looking for a har alternative? If you're comparing har with other mcp server tools, these 6 projects are the closest alternatives on Agent Skills Hub — ranked by topic overlap, star count, and community traction.
Lightweight agent orchestration dev tool that lives in your repo
Server-enforced workflow discipline for AI agents. An MCP server providing persistent work items, dependency g
MCP Fusion - The framework for AI-native MCP servers.
Git worktree-based task manager with integrated terminals for CLI coding agents
Local-first agent platform: board + role agents + Claude Code runs, all on your own Mac
Local-first coordination for human and agent work: durable work, decisions, dispatches, evidence, and prompt-f
Explore other popular mcp server tools:
har is HAR: open agent harness (CLI + MCP) for coding agents. Isolated worktrees, deterministic verify, software factory workflows for Claude Code, Cursor, and Codex.. It is categorized as a MCP Server with 90 GitHub stars.
har is primarily written in TypeScript. It covers topics such as agents, ai-agents, ai-coding.
You can find installation instructions and usage details in the har GitHub repository at github.com/os-factory/har. The project has 90 stars and 11 forks, indicating an active community.
har is released under the Apache-2.0 license, making it free to use and modify according to the license terms.
The top alternatives to har on Agent Skills Hub include sudocode, task-orchestrator, mcp-fusion. Each offers a different approach to the same problem space — compare them side-by-side by stars, quality score, and community activity.
Grades come from a rule-based scan built on the SlowMist agent-security taxonomy, covering 11 red-flag categories including credential harvesting, data exfiltration, and curl | sh installers. It is a first-layer scan, not a manual audit — we say so rather than overstate it.
The scale of the problem is documented independently: Liu et al. (2026), in a study of 31,132 agent skills, report that 26.1% contain security vulnerabilities. Our own full-catalog census is published as a citable open dataset.
Sources & who's responsible: