Qwen Code

FullL2 MCP

Alibaba

Gemini CLI fork — QWEN.md instruction file

CommandsRulesSkillsHooksMCPrules: .md

CRASH Capability Matrix

CRASH = Commands · Rules · Agents · Skills · Hooks + MCP. Engine level: L2 — MCP Native

PrimitiveStatus
CommandscommandsCommands
RulesrulesRules
SkillsskillsSkills
HookshooksHooks
MCPmcpMCP

Folder Hierarchy

Files gald3r writes and the directories it expects for Qwen Code. The .qwen/ tree is the primary gald3r surface.

project-root/
.qwen/
└── settings.json   ← hooks + MCP config
QWEN.md             ← primary instruction file (not AGENTS.md by default)

Key Setup Notes

  • Alibaba Qwen coding agent — architecturally a Gemini CLI fork
  • Primary instruction file is QWEN.md (not AGENTS.md)
  • Hooks and MCP via .qwen/settings.json
  • Subagents and Agent Skills (SKILL.md) natively supported

Known Gaps

Areas where gald3r's native capability is limited or absent on Qwen Code. The engine compensates where possible via L2 — MCP Native.

  • Reads QWEN.md — not AGENTS.md (by default)

Install gald3r on Qwen Code

Run the gald3r installer and get the full CRASH system on Qwen Code in under 5 minutes. Persistent memory, 89+ commands, and adversarial quality gates.

All platforms
35 platforms supported