Agent delegation not working
"models": {
"mode": "merge",
"providers": {
"ollama-m3": {
"baseUrl": "http://192.168.1.170:11434/v1",
"apiKey": "",
"api": "ollama",
"models": [
{
"id": "gpt-oss:20b",
"name": "gpt-oss:20b",
"reasoning": true,
"input": [
"text"
],
"contextWindow": 16384,
"maxTokens": 32000
}
]
},
"ollama-3080": {
"baseUrl": "http://192.168.1.76:11434/v1",
"apiKey": "",
"api": "ollama",
"models": [
{
"id": "qwen3.5:27b",
"name": "qwen3.5-27b",
"reasoning": false,
"input": [
"text"
],
"contextWindow": 32768,
"maxTokens": 32000
}
]
}
}
},
"agents": {
"defaults": {
"model": {
"primary": "ollama-m3/gpt-oss:20b",
"fallbacks": []
},
"models": {
"ollama-m3/gpt-oss:20b": {},
"ollama-3080/qwen3.5:27b": {}
},
"workspace": "~/.openclaw/workspace",
"compaction": {
"mode": "safeguard"
},
"maxConcurrent": 1,
"subagents": {
"maxConcurrent": 2
}
},
"list": [
{
"id": "main",
"subagents": {
"allowAgents": [
"coder"
]
}
},
{
"id": "coder",
"workspace": "~/.openclaw/workspace-coder",
"model": {
"primary": "ollama-3080/qwen3.5:27b"
}
}
]
},
"tools": {
"profile": "full"
},
"bindings": [],
gpt-oss-20b responds but never delegates to qwen3.5, both models respond via curl independently, SOUL.md and AGENT.md are explicit to send all coding requests to qwen3.5
"mode": "merge",
"providers": {
"ollama-m3": {
"baseUrl": "http://192.168.1.170:11434/v1",
"apiKey": "",
"api": "ollama",
"models": [
{
"id": "gpt-oss:20b",
"name": "gpt-oss:20b",
"reasoning": true,
"input": [
"text"
],
"contextWindow": 16384,
"maxTokens": 32000
}
]
},
"ollama-3080": {
"baseUrl": "http://192.168.1.76:11434/v1",
"apiKey": "",
"api": "ollama",
"models": [
{
"id": "qwen3.5:27b",
"name": "qwen3.5-27b",
"reasoning": false,
"input": [
"text"
],
"contextWindow": 32768,
"maxTokens": 32000
}
]
}
}
},
"agents": {
"defaults": {
"model": {
"primary": "ollama-m3/gpt-oss:20b",
"fallbacks": []
},
"models": {
"ollama-m3/gpt-oss:20b": {},
"ollama-3080/qwen3.5:27b": {}
},
"workspace": "~/.openclaw/workspace",
"compaction": {
"mode": "safeguard"
},
"maxConcurrent": 1,
"subagents": {
"maxConcurrent": 2
}
},
"list": [
{
"id": "main",
"subagents": {
"allowAgents": [
"coder"
]
}
},
{
"id": "coder",
"workspace": "~/.openclaw/workspace-coder",
"model": {
"primary": "ollama-3080/qwen3.5:27b"
}
}
]
},
"tools": {
"profile": "full"
},
"bindings": [],
gpt-oss-20b responds but never delegates to qwen3.5, both models respond via curl independently, SOUL.md and AGENT.md are explicit to send all coding requests to qwen3.5
The AI that actually does things. Emails, calendar, home automation - all from your favorite chat app. New shell, ready to help. The lobster way.
134,426Members
Resources
Similar Threads
Was this page helpful?
