Skillget

mstar-harness

agent
Guvenlik Denetimi
Uyari
Health Uyari
  • License — License: MIT
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Low visibility — Only 5 GitHub stars
Code Gecti
  • Code scan — Scanned 3 files during light audit, no dangerous patterns found
Permissions Gecti
  • Permissions — No dangerous permissions requested
Purpose
This project provides a multi-agent harness framework called "Morning Star" designed to coordinate multi-role workflows. It acts as a shared configuration and skill layer for developer tools like OpenCode and Cursor.

Security Assessment
Overall risk: Low. The codebase was scanned and contains no dangerous patterns or hardcoded secrets. The tool does not request dangerous permissions or execute hidden shell commands beyond standard setup routines. However, the installation relies heavily on using symlinks and executing `rm -rf` commands to replace existing configuration directories. While standard for developer setups, you should carefully review these commands before running them to avoid accidental data loss. Additionally, the configuration requires you to supply your own API keys or secrets via local environment variables or files, which keeps credentials out of the code but requires secure handling on your end.

Quality Assessment
The framework is licensed under the permissive MIT standard. It is actively maintained, with recent repository pushes occurring within the last day. On the downside, the project currently has very low community visibility with only five GitHub stars, meaning it has not been extensively battle-tested by a wide audience.

Verdict
Safe to use, but proceed with standard caution regarding terminal commands and evaluate whether you are comfortable adopting a low-visibility, early-stage framework.
SUMMARY

A skill-driven Harness Engineering framework for Opencode/Cursor

README.md

Morning Star — Code Agent Harness Framework

English / 中文

GitHub · Issues


This repository provides the Morning Star multi-agent code harness framework.

Core value:

  • Start a usable multi-role workflow quickly
  • Run with unified mstar-* skills instead of scattered rules
  • Reuse one core process across OpenCode and Cursor

Quick Start (Recommended)

Use a single source checkout and install via symlinks:

  1. Clone repository to a stable source path:
    • git clone https://github.com/btspoony/mstar-harness.git ~/.mstar-harness
  2. Install for OpenCode (symlink config root):
    • ln -s ~/.mstar-harness ~/.config/opencode
    • If target exists, recreate safely:
    • rm -rf ~/.config/opencode && ln -s ~/.mstar-harness ~/.config/opencode
  3. Install for Cursor local plugin (symlink plugin root):
    • mkdir -p ~/.cursor/plugins/local
    • ln -s ~/.mstar-harness ~/.cursor/plugins/local/mstar-harness
    • If link exists, recreate safely:
    • rm -f ~/.cursor/plugins/local/mstar-harness && ln -s ~/.mstar-harness ~/.cursor/plugins/local/mstar-harness
    • Or force update:
    • ln -sfn ~/.mstar-harness ~/.cursor/plugins/local/mstar-harness
  4. Create local config:
    • cp ~/.config/opencode/opencode.example.json ~/.config/opencode/opencode.json
  5. Configure secrets via {env:...} or {file:...} placeholders
  6. Restart OpenCode / Cursor session and verify entry files are readable

If you prefer incremental adoption (without replacing the whole directory), merge these first:

  • AGENTS.md
  • agents/
  • skills/mstar-*/
  • .opencode/skills/mstar-host/
  • .cursor-plugin/skills/mstar-host/

Cursor Local Plugin Install (Symlink)

If you already cloned to ~/.mstar-harness, install Cursor plugin via:

  1. Create local plugin folder:
    • mkdir -p ~/.cursor/plugins/local
  2. Create symlink:
    • ln -s ~/.mstar-harness ~/.cursor/plugins/local/mstar-harness
    • If the link already exists, recreate it safely:
    • rm -f ~/.cursor/plugins/local/mstar-harness && ln -s ~/.mstar-harness ~/.cursor/plugins/local/mstar-harness
    • Or use force update:
    • ln -sfn ~/.mstar-harness ~/.cursor/plugins/local/mstar-harness
  3. Restart Cursor or run Developer: Reload Window.
  4. Verify plugin components are loaded (rules, skills, agents).

Host Entry (OpenCode vs Cursor)

Host First thing to do
OpenCode Usually auto-injects AGENTS.md; then use .opencode/skills/mstar-host/SKILL.md for host-specific behavior
Cursor Read AGENTS.md manually first, then .cursor-plugin/skills/mstar-host/SKILL.md

Recommended sequence for both hosts:

  1. AGENTS.md
  2. Current host mstar-host skill
  3. skills/mstar-roles/SKILL.md
  4. Target role Role reference

Role and Skill Overview

Roles

Agent ID Role Responsibility
project-manager Project Manager Routing, assignment, phase progression
product-manager Product Manager Requirements and product-facing docs
architect Architect Architecture and technical contracts
fullstack-dev / fullstack-dev-2 Fullstack Dev Backend-led implementation / second parallel track
frontend-dev Frontend Dev UI, interaction, frontend performance
qa-engineer QA Testing and acceptance validation
qc-specialist* QC Trio Code quality gate (architecture/security/performance)
ops-engineer Ops Deployment, monitoring, infrastructure
market-expert Market Expert Market and user research
prompt-engineer Prompt Engineer Prompt / skill / rule optimization

Core Skills

Skill Purpose
mstar-harness-core Global entry, state machine, gates, invariants
mstar-plan-conventions Unified plan/status/residual conventions
mstar-review-qc QC review baseline and report template
mstar-coding-behavior Cross-role coding behavior baseline
mstar-superpowers-align Alignment and conflict handling with Superpowers
mstar-roles Role prompt bus (role bodies in references/)
mstar-host (per host) Host-specific capabilities (OpenCode / Cursor)

Morning Star load order: In any session or task, read skills/mstar-harness-core/SKILL.md before any other skills/mstar-*/SKILL.md. Each non-core skill begins with a Load order section that repeats this; on conflict, mstar-harness-core wins. See mstar-harness-core/SKILL.md →「与其它 mstar-* skill 的加载契约」.

Cursor-only maint: PM routing scenario regression + Routing Eval Report live in .cursor/skills/mstar-routing-eval/ (maint-only), while runtime host skills live in .cursor-plugin/skills/. See .cursor/rules/repo-maintenance.mdc.

Common Flows (Short)

I want to start a task now

  1. Read AGENTS.md
  2. Let project-manager establish context and route
  3. Execution roles work via mstar-roles + relevant skills
  4. Close through QC/QA gates

I just need rule entry points

  • Global entry: AGENTS.md
  • Core workflow: skills/mstar-harness-core/SKILL.md
  • Role hub: skills/mstar-roles/SKILL.md

Deeper Docs

  • AGENTS.md: global entry and index
  • skills/mstar-harness-core/: core execution rules
  • skills/mstar-roles/: role content and parameterization
  • skills/mstar-plan-conventions/: plan and state conventions

Maintainer workflows, cross-file sync rules, and lint/maintenance checklists are intentionally kept in maintainer rule docs, not in this user-facing README.

License

This project is licensed under MIT. See LICENSE.

Yorumlar (0)

Sonuc bulunamadi