chore(sync): mirror docs from openclaw/openclaw@0603c2327d
This commit is contained in:
parent
282f46f264
commit
653ca8d0f9
@ -1,5 +1,5 @@
|
||||
{
|
||||
"repository": "openclaw/openclaw",
|
||||
"sha": "a89fe0f5a0ce8b48e3c388aef01da71eafbeca34",
|
||||
"syncedAt": "2026-04-30T03:15:17.994Z"
|
||||
"sha": "0603c2327d3e09540b4488596aa3062c6acfc550",
|
||||
"syncedAt": "2026-04-30T04:04:44.074Z"
|
||||
}
|
||||
|
||||
@ -1,4 +1,4 @@
|
||||
7bf720f6d9040c53323553b1bd351f688137c6b352c4cf2acfd7f7d252644b38 config-baseline.json
|
||||
cb32b51492306179b4537514b0650ab24e2f5f8f6c2eda92154cb1420a11e560 config-baseline.json
|
||||
ab9a004ec78ed51e646be29eb10aa6700de1d47fee77331a85ca5e2cd15b6e93 config-baseline.core.json
|
||||
92712871defa92eeda8161b516db85574681f2b70678b940508a808b987aeae2 config-baseline.channel.json
|
||||
c4231c2194206547af8ad94342dc00aadb734f43cb49cc79d4c46bdbb80c3f95 config-baseline.plugin.json
|
||||
ede8b3d9bd7848a09abfcd9fa4f007d289d05742f66b0e38ef459da6dbf40897 config-baseline.plugin.json
|
||||
|
||||
@ -1,2 +1,2 @@
|
||||
9262e43a0171f4c38a0590fe36d80b13382beee4a1be1ae2ffb109ace5f37b31 plugin-sdk-api-baseline.json
|
||||
c7385f6584052938fe9cce00ae5f9d90cb3b08f943cc82122235f18a1213439b plugin-sdk-api-baseline.jsonl
|
||||
dd840b7c222ca003aa5336aabff8a126e3e254474941ddab93165e0e44944ffa plugin-sdk-api-baseline.json
|
||||
443878722940029e4ae5220f3c23ffc321559b73848f6a7a3f4cab98c076924e plugin-sdk-api-baseline.jsonl
|
||||
|
||||
@ -202,6 +202,12 @@ Dangerous or privacy-heavy commands such as `camera.snap`, `camera.clip`, and
|
||||
`gateway.nodes.allowCommands`. `gateway.nodes.denyCommands` always wins over
|
||||
defaults and extra allowlist entries.
|
||||
|
||||
Plugin-owned node commands can add a Gateway node-invoke policy. That policy
|
||||
runs after the allowlist check and before forwarding to the node, so raw
|
||||
`node.invoke`, CLI helpers, and dedicated agent tools share the same plugin
|
||||
permission boundary. Dangerous plugin node commands still require explicit
|
||||
`gateway.nodes.allowCommands` opt-in.
|
||||
|
||||
After a node changes its declared command list, reject the old device pairing
|
||||
and approve the new request so the gateway stores the updated command snapshot.
|
||||
|
||||
|
||||
@ -178,7 +178,9 @@ Provider and channel execution paths must use the active runtime config snapshot
|
||||
});
|
||||
```
|
||||
|
||||
Inside the Gateway this runtime is in-process. In plugin CLI commands it calls the configured Gateway over RPC, so commands such as `openclaw googlemeet recover-tab` can inspect paired nodes from the terminal. Node commands still go through normal Gateway node pairing, command allowlists, and node-local command handling.
|
||||
Inside the Gateway this runtime is in-process. In plugin CLI commands it calls the configured Gateway over RPC, so commands such as `openclaw googlemeet recover-tab` can inspect paired nodes from the terminal. Node commands still go through normal Gateway node pairing, command allowlists, plugin node-invoke policies, and node-local command handling.
|
||||
|
||||
Plugins that expose dangerous node-host commands should register a node-invoke policy with `api.registerNodeInvokePolicy(...)`. The policy runs in the Gateway after command allowlist checks and before the command is forwarded to the node, so direct `node.invoke` calls and higher-level plugin tools share the same enforcement path.
|
||||
|
||||
</Accordion>
|
||||
<Accordion title="api.runtime.tasks.managedFlows">
|
||||
|
||||
@ -619,6 +619,7 @@ Active Memory 會透過已設定記憶體 Plugin 的召回管線運作,因此
|
||||
- 觀察 Gateway 記錄中的 `active-memory: ... start|done`、`memory sync failed (search-bootstrap)`,或提供者嵌入錯誤。
|
||||
- 執行 `openclaw memory status --deep`,檢查記憶體搜尋後端與索引健康狀態。
|
||||
- 如果你使用 `ollama`,請確認嵌入模型已安裝(`ollama list`)。
|
||||
|
||||
</Accordion>
|
||||
</AccordionGroup>
|
||||
|
||||
|
||||
@ -314,6 +314,7 @@ openclaw models list
|
||||
接著是 `AWS_ACCESS_KEY_ID` + `AWS_SECRET_ACCESS_KEY`,再來是 `AWS_PROFILE`,最後是預設 AWS SDK 鏈。
|
||||
- 推理支援取決於模型;請查看 Bedrock 模型卡以了解目前功能。
|
||||
- 如果你偏好受管理的金鑰流程,也可以在 Bedrock 前方放置 OpenAI 相容的代理,並將其改為設定成 OpenAI 提供者。
|
||||
|
||||
</Accordion>
|
||||
</AccordionGroup>
|
||||
|
||||
|
||||
@ -214,6 +214,7 @@ LiteLLM 也可以透過 OpenAI 相容的
|
||||
沒有 `service_tier`、沒有 Responses `store`、沒有提示快取提示,也沒有
|
||||
OpenAI reasoning 相容酬載塑形
|
||||
- 自訂 LiteLLM base URL 不會注入隱藏的 OpenClaw 歸因標頭(`originator`、`version`、`User-Agent`)
|
||||
|
||||
</Accordion>
|
||||
</AccordionGroup>
|
||||
|
||||
|
||||
@ -393,6 +393,7 @@ OpenClaw 使用 xAI 的 REST 圖片/影片/TTS/STT API 進行媒體生成、語
|
||||
- 一般 xAI 供應商路徑不支援 `grok-4.20-multi-agent-experimental-beta-0304`,因為它需要與標準 OpenClaw xAI 傳輸不同的上游 API 介面。
|
||||
- xAI Realtime 語音尚未註冊為 OpenClaw 供應商。它需要與批次 STT 或串流轉錄不同的雙向語音工作階段合約。
|
||||
- xAI 圖片 `quality`、圖片 `mask`,以及額外的僅原生支援長寬比,在共用 `image_generate` 工具具備相對應的跨供應商控制之前不會公開。
|
||||
|
||||
</Accordion>
|
||||
|
||||
<Accordion title="進階備註">
|
||||
@ -402,6 +403,7 @@ OpenClaw 使用 xAI 的 REST 圖片/影片/TTS/STT API 進行媒體生成、語
|
||||
- `web_search`、`x_search` 和 `code_execution` 會作為 OpenClaw 工具公開。OpenClaw 會在每個工具請求內啟用所需的特定 xAI 內建功能,而不是將所有原生工具附加到每個聊天輪次。
|
||||
- `x_search` 和 `code_execution` 由內建的 xAI Plugin 擁有,而不是硬編碼到核心模型執行階段。
|
||||
- `code_execution` 是遠端 xAI 沙箱執行,不是本機 [`exec`](/zh-TW/tools/exec)。
|
||||
|
||||
</Accordion>
|
||||
</AccordionGroup>
|
||||
|
||||
|
||||
@ -72,6 +72,7 @@ OpenClaw 會從這些來源載入技能,**優先順序由高到低**:
|
||||
- 設定 `agents.list[].skills: []` 表示沒有技能。
|
||||
- 非空的 `agents.list[].skills` 清單就是該代理的**最終**集合,不會與預設值合併。
|
||||
- 有效允許清單會套用於提示建構、技能斜線命令探索、沙箱同步與技能快照。
|
||||
|
||||
</Accordion>
|
||||
</AccordionGroup>
|
||||
|
||||
|
||||
Loading…
Reference in New Issue
Block a user