opencode 实战经验分享（二）解决使用 CPA 的模型无法正常压缩上下文的问题
之前分享了一个 opencode 的配置，大家也给我提了很多意见，感谢大家！！！
但是我在使用 opencode 的时候当上下文快满的时候自动进行上下文压缩，但是会报错，然后我尝试了手动压缩 /compact 指令，我发现了一个很有意思的现象，当我不使用工具调用的时候可以正常压缩，但是我只要使用工具调用就报错：Bad Request: Improperly formed request. 所以我就猜测是因为思考块里面带了工具调用，致使 CPA 不支持，所以我就开始改 CPA，修复逻辑如下
然后吭哧吭哧改了两天，调了很多次，最后可以正常压缩了，但是工具调用的时候一直循环，无奈宣告失败。。。
但是，我在查文档的时候发现压缩上下文的也是一个子代理，默认使用的是主模型，那既然这样，我换一个模型不就行了，直接开干，因为我设置了 google 的 api 所以我用的是 google 模型。
正常工作，果然，opencode 的高自定义程度名不虚传，好玩爱玩，如果有佬有更好的方法，欢迎分享，在此献上我的 opencode.json 文件，大家可以看看：
{
"$schema": "https://opencode.ai/config.json",
"model": "cpa-claude/kiro-claude-opus-4-5-agentic(high)",
"agent": {
"compaction": {
"model": "google/antigravity-gemini-3-pro-low"
}
},
"plugin": [
"oh-my-opencode",
"opencode-antigravity-auth@1.2.8",
"opencode-openai-codex-auth@4.3.0"
],
"provider": {
"cpa-claude": {
"npm": "@ai-sdk/anthropic",
"name": "cpa-claude",
"options": {
"baseURL": "http://127.0.0.1:8317/v1"},
"models": {
"gemini-claude-opus-4-5-thinking(high)": {
"name": "gemini-claude-opus-4-5-thinking(high)",
"limit": {
"context": 204800,
"output": 65535
},
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
},
"options": {
"thinking": {
"type": "enabled",
"budgetTokens": 16000
}
}
},
"gemini-claude-sonnet-4-5-thinking(high)": {
"name": "gemini-claude-sonnet-4-5-thinking(high)",
"limit": {
"context": 204800,
"output": 65535
},
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
},
"options": {
"thinking": {
"type": "enabled",
"budgetTokens": 16000
}
}
},
"kiro-claude-sonnet-4-5-agentic": {
"name": "claude-sonnet-4-5-nothinking",
"limit": {
"context": 204800,
"output": 65535
},
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
}
},
"kiro-claude-sonnet-4-5-agentic(high)": {
"name": "claude-sonnet-4-5",
"limit": {
"context": 204800,
"output": 65535
},
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
},
"options": {
"thinking": {
"type": "enabled",
"budgetTokens": 16000
}
}
},
"kiro-claude-opus-4-5-agentic(high)": {
"name": "claude-opus-4-5",
"limit": {
"context": 204800,
"output": 65535
},
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
},
"options": {
"thinking": {
"type": "enabled",
"budgetTokens": 16000
}
}
}
}
},
"cpa-gemini": {
"name": "cpa-gemini",
"options": {
"baseURL": "http://127.0.0.1:8317/v1"
},
"models": {
"gemini-3-pro-preview": {
"name": "gemini-3-pro-preview",
"limit": {
"context": 1048576,
"output": 65535
},
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
},
"options": {
"thinking": {
"type": "enabled",
"budgetTokens": 16000
}
}
},
"gemini-3-flash-preview": {
"name": "gemini-3-flash-preview",
"limit": {
"context": 1048576,
"output": 65536
},
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
},
"options": {
"thinking": {
"type": "enabled",
"budgetTokens": 16000
}
}
}
}
},
"google": {
"name": "Google",
"models": {
"antigravity-gemini-3-pro-high": {
"name": "Gemini 3 Pro High (Antigravity)",
"thinking": true,
"attachment": true,
"limit": {
"context": 1048576,
"output": 65535
},
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
}
},
"antigravity-gemini-3-pro-low": {
"name": "Gemini 3 Pro Low (Antigravity)",
"thinking": true,
"attachment": true,
"limit": {
"context": 1048576,
"output": 65535
},
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
}
},
"antigravity-gemini-3-flash": {
"name": "Gemini 3 Flash (Antigravity)",
"attachment": true,
"limit": {
"context": 1048576,
"output": 65536
},
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
}
}
}
},
"openai": {
"name": "OpenAI",
"options": {
"reasoningEffort": "medium ",
"reasoningSummary": "auto",
"textVerbosity": "medium ",
"include": [
"reasoning.encrypted_content"
],
"store": false
},
"models": {
"gpt-5.2": {
"name": "GPT 5.2 (OAuth)",
"limit": {
"context": 272000,
"output": 128000
},
"modalities": {
"input": [
"text",
"image"
],
"output": [
"text"
]
},
"variants": {
"none": {
"reasoningEffort": "none",
"reasoningSummary": "auto",
"textVerbosity": "medium "
},
"low": {
"reasoningEffort": "low",
"reasoningSummary": "auto",
"textVerbosity": "medium "
},
"medium ": {
"reasoningEffort": "medium ",
"reasoningSummary": "auto",
"textVerbosity": "medium "
},
"high": {
"reasoningEffort": "high",
"reasoningSummary": "detailed",
"textVerbosity": "medium "
},
"xhigh": {
"reasoningEffort": "xhigh",
"reasoningSummary": "detailed",
"textVerbosity": "medium "
}
}
},
"gpt-5.2-codex": {
"name": "GPT 5.2 Codex (OAuth)",
"limit": {
"context": 272000,
"output": 128000
},
"modalities": {
"input": [
"text",
"image"
],
"output": [
"text"
]
},
"variants": {
"low": {
"reasoningEffort": "low",
"reasoningSummary": "auto",
"textVerbosity": "medium "
},
"medium ": {
"reasoningEffort": "medium ",
"reasoningSummary": "auto",
"textVerbosity": "medium "
},
"high": {
"reasoningEffort": "high",
"reasoningSummary": "detailed",
"textVerbosity": "medium "
},
"xhigh": {
"reasoningEffort": "xhigh",
"reasoningSummary": "detailed",
"textVerbosity": "medium "
}
}
},
"gpt-5.1-codex-max": {
"name": "GPT 5.1 Codex Max (OAuth)",
"limit": {
"context": 272000,
"output": 128000
},
"modalities": {
"input": [
"text",
"image"
],
"output": [
"text"
]
},
"variants": {
"low": {
"reasoningEffort": "low",
"reasoningSummary": "detailed",
"textVerbosity": "medium "
},
"medium ": {
"reasoningEffort": "medium ",
"reasoningSummary": "detailed",
"textVerbosity": "medium "
},
"high": {
"reasoningEffort": "high",
"reasoningSummary": "detailed",
"textVerbosity": "medium "
},
"xhigh": {
"reasoningEffort": "xhigh",
"reasoningSummary": "detailed",
"textVerbosity": "medium "
}
}
}
}
}
}
}
除了使用 google/antigravity-gemini-3-pro-low 还可以使用 opencode 提供的免费模型，比如 GLM4.7。
当然，这个只是一个曲线救国的方法，如果有佬有更好的方法，欢迎分享！！！
配置问题可以看看之前那个帖子