mknormal

Open Swarmforce — DevOps skill for Claude Code

DevOps community

Agentic CI/CD gate: reviews your diff, fixes confirmed findings in isolated sandboxes, re-reviews, and requires your own tests to pass before a PR publishes.

How to install Open Swarmforce

This entry records only its repository, not the path inside it, so there is no exact command to give. Open mknormal/open_swarmforce and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Open Swarmforce does

Agentic CI/CD gate: reviews your diff, fixes confirmed findings in isolated sandboxes, re-reviews, and requires your own tests to pass before a PR publishes. MIT fork of OpenSwarm with a rebranded dashboard. Works with Codex/GPT, OpenRouter, local models, Claude.

Alternatives in DevOps

  • 16 Testing — Prompt 16: Add Test Infrastructure & Smoke Tests 2.3k ★
  • Skill — Web extraction engine with antibot bypass 549 ★
  • Zhikuncode — Codex/Claude Code/Cursor的开源增强版,专注一句话实现复杂长程任务(教育、编程、办公、生活、娱乐、游戏) 473 ★

README

SWARMFORGE

SWARMFORGE

[![license](https://img.shields.io/badge/license-MIT-green.svg)](LICENSE) [![fork of OpenSwarm](https://img.shields.io/badge/fork_of-OpenSwarm-4F46E5)](https://github.com/Intrect-io/OpenSwarm) [![SWE-bench Lite](https://img.shields.io/badge/SWE--bench_Lite-hybrid_3%2F3_resolved-2ea44f)](benchmarks/RUBRIC.md)

Agentic CI/CD gate — review, fix, re-review, and deterministic verification for pull requests and local changes

**SWARMFORGE is a frontend-focused fork of [OpenSwarm](https://github.com/Intrect-io/OpenSwarm).** The upstream engine — CLI, daemon, agent pipeline, provider adapters, task state, and execution infrastructure — is kept intact; v1 changes only user-facing branding, dashboard presentation, metadata, and documentation. Upstream is MIT-licensed and its notices are preserved in [`LICENSE`](LICENSE) and [`NOTICE.md`](NOTICE.md); this fork is not affiliated with or endorsed by the OpenSwarm project.


SWARMFORGE is an **agentic CI/CD gate** for the change path that matters: it reviews a committed or working-tree diff, can group and fix confirmed findings in isolated sandboxes, re-reviews the result, then requires deterministic repository checks before it publishes a PR. It runs with **OpenAI Codex/GPT**, **any OpenRouter model**, **local open-source models** (Ollama, LM Studio), or **Claude Code** (`claude -p`, opt-in) — with cost-aware routing measured on an L0–L6 benchmark ladder.

Use it locally before a commit or from the included GitHub Action on a pull request. The Linear-backed autonomous daemon, notifications, and long-term repository memory remain available as an optional orchestration layer; they are not required to use the merge gate.

**Verified on real GitHub issues**: the agentic harness solves SWE-bench Lite instances graded by the official harness. Hybrid mode — a frontier model diagnoses read-only,