@huaweidevtools/aishell-dsh-llm 0.0.0-stage → 0.0.1-beta.0

package/LICENSE

@@ -0,0 +1,201 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined in Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is derived from (or based upon) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright 2024 Huawei Cloud AI Shell
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+   implied. See the License for the specific language governing
+   permissions and limitations under the License.

package/README.md

@@ -1,3 +1,66 @@
-# Temporary Holding Version
+# @huaweidevtools/aishell-dsh-llm
 
-This version is a temporary placeholder for this package. An operational version to replace this has been submitted for review and is awaiting a staged release.
+OpenAI-compatible LLM adapter for DSH. Registers the `aishell` provider route on `ctx.llm`, routing model calls through the aishell gateway with direct AK/SK signing.
+
+## How it works
+
+```
+User selects aishell/<model>
+  → GatewayAdapter.stream(options)
+
+  Every request is signed directly with AK/SK (SDK-HMAC-SHA256):
+    → credential store reads HW_ACCESS_KEY + HW_SECRET_KEY (+ optional HW_SECURITY_TOKEN)
+    → sign(POST, host, /chat/completions, query, body) → signed headers
+    → POST {baseUrl}/chat/completions  (signed headers, OpenAI wire format)
+    → parseSse → translate → StreamChunks
+
+  Model listing:
+    → GET {baseUrl}/models  (AK/SK signed)
+    → returns OpenAI-compatible model list
+```
+
+No intermediate API-key exchange or config fetch — every request is signed with the user's AK/SK. The credential store (OS keyring with fingerprint + memory fallback) holds the AK/SK; this adapter only reads them.
+
+## What it does NOT touch
+
+- `ctx.credentials` — DSH's own credential provider stays `dsh-credentials-local`. User-configured models and `.credentials.yaml` keep working.
+- `llm-deepseek` / other adapters — user's existing providers are unchanged.
+- No tools, commands, or session behavior is modified.
+
+## Install
+
+```sh
+# Standalone (LLM adapter only — no credentials UI):
+dsh plugin --profile <name> add link:<repo>/packages/aishell-llm
+
+# Or via the aishell-base bundle (full stack):
+dsh plugin --profile <name> add @huaweidevtools/aishell-dsh-base
+```
+
+## Config
+
+| Field | Type | Default | Description |
+|---|---|---|---|
+| `baseUrl` | `string?` | `https://api.aishell.openbuildlab.com` | Chat API base URL. `/chat/completions` and `/models` are appended. |
+| `keyringService` | `string?` | `aishell-dsh` | Credential store service name. |
+
+## Storing credentials
+
+**Web UI (recommended):** Install `aishell-credentials` (or the `aishell-base` bundle) — the Settings page gets an "AI Shell" section with AK/SK/SecurityToken input fields. Values are transferred to the credential store.
+
+**CLI fallback:**
+
+```sh
+node --input-type=module -e "
+import { createCredentialsStore } from '@huaweidevtools/aishell-dsh-credentials-store';
+const s = createCredentialsStore();
+s.set('HW_ACCESS_KEY', 'your-ak');
+s.set('HW_SECRET_KEY', 'your-sk');
+// Optional: temporary security token (STS)
+s.set('HW_SECURITY_TOKEN', 'your-token');
+"
+```
+
+## License
+
+Apache License 2.0.

package/cordis.patch.yml

@@ -0,0 +1,7 @@
+# aishell-llm bundle patch: inserts the aishell LLM gateway adapter.
+# Install standalone: `dsh plugin --profile <name> add link:<repo>/packages/aishell-llm`
+# The single row mounts the host half (LLM adapter reading AK/SK from the
+# credential store). Credentials are managed by aishell-credentials.
+- insert:
+    - id: aishell-llm
+      name: '@huaweidevtools/aishell-dsh-llm'

package/lib/index.js

@@ -0,0 +1,423 @@
+import { createCredentialsStore } from "@huaweidevtools/aishell-dsh-credentials-store";
+import { LlmAdapter, LlmError } from "@deepseek-ai/dsh-llm";
+import { brandString } from "@deepseek-ai/dsh-brand";
+import { EventSourceParserStream } from "eventsource-parser/stream";
+import { sign } from "@huaweidevtools/aishell-dsh-credentials-store/signer";
+
+//#region src/serialize.ts
+/** Build the POST body for `/chat/completions`. */
+function serializeRequest(options) {
+	const messages = [];
+	for (const message of options.messages) if (message.role === "user") {
+		const textBlocks = message.content.filter((block) => block.type === "text" || block.type === "reasoning");
+		messages.push({
+			role: "user",
+			content: textBlocks.length > 0 ? blocksToText(textBlocks) : ""
+		});
+	} else if (message.role === "tool") messages.push({
+		role: "tool",
+		tool_call_id: String(message.toolCallId),
+		content: blocksToText(message.content)
+	});
+	else if (message.role === "developer") {
+		const text = blocksToText(message.content);
+		if (text) messages.push({
+			role: "system",
+			content: text
+		});
+	} else if (message.role === "system") messages.push({
+		role: "system",
+		content: blocksToText(message.content)
+	});
+	else {
+		const toolCalls = message.content.filter(isToolCallBlock).map(serializeAssistantToolCall);
+		const assistantText = blocksToText(message.content.filter((block) => !isToolCallBlock(block)));
+		messages.push({
+			role: "assistant",
+			content: assistantText || null,
+			...toolCalls.length > 0 ? { tool_calls: toolCalls } : {}
+		});
+	}
+	return {
+		model: options.model,
+		messages,
+		stream: true,
+		stream_options: { include_usage: true },
+		...options.tools && options.tools.length > 0 ? { tools: options.tools.map(serializeTool) } : {},
+		...options.temperature !== void 0 ? { temperature: options.temperature } : {},
+		...options.maxTokens !== void 0 ? { max_tokens: options.maxTokens } : {},
+		...options.stop !== void 0 ? { stop: options.stop } : {}
+	};
+}
+/** Type guard for tool-call blocks. */
+function isToolCallBlock(block) {
+	return block.type === "tool-call";
+}
+function blocksToText(blocks) {
+	return blocks.filter((block) => block.type === "text" || block.type === "reasoning").map((block) => block.text ?? "").join("");
+}
+function serializeAssistantToolCall(block) {
+	return {
+		id: String(block.id),
+		type: "function",
+		function: {
+			name: block.name,
+			arguments: block.arguments
+		}
+	};
+}
+function serializeTool(tool) {
+	return {
+		type: "function",
+		function: {
+			name: tool.name,
+			description: tool.description,
+			parameters: tool.parameters
+		}
+	};
+}
+
+//#endregion
+//#region src/sse.ts
+const DONE = "[DONE]";
+async function* parseSse(stream) {
+	const events = stream.pipeThrough(new TextDecoderStream()).pipeThrough(new EventSourceParserStream());
+	for await (const { data } of events) {
+		yield data;
+		if (data === DONE) return;
+	}
+	throw new LlmError("SSE stream ended without [DONE]", "STREAM_CLOSED");
+}
+
+//#endregion
+//#region src/translate.ts
+/** Map the wire finish_reason vocabulary to the harness FinishReason. */
+function mapFinishReason(reason) {
+	switch (reason) {
+		case "stop": return { kind: "stop" };
+		case "tool_calls": return { kind: "tool-calls" };
+		case "length": return { kind: "max-tokens" };
+		default: return {
+			kind: "error",
+			failure: {
+				message: `model stopped: ${reason ?? "unknown"}`,
+				code: (reason ?? "unknown").toUpperCase()
+			}
+		};
+	}
+}
+/** Map wire usage fields to disjoint harness counts. */
+function mapUsage(usage) {
+	const cacheRead = usage.prompt_tokens_details?.cached_tokens;
+	const reasoning = usage.completion_tokens_details?.reasoning_tokens;
+	const combined = usage.prompt_tokens + usage.completion_tokens;
+	const hasExactTotal = Number.isSafeInteger(usage.prompt_tokens) && usage.prompt_tokens >= 0 && Number.isSafeInteger(usage.completion_tokens) && usage.completion_tokens >= 0 && Number.isSafeInteger(combined) && (usage.total_tokens === void 0 || usage.total_tokens === combined);
+	return {
+		inputTokens: usage.prompt_tokens - (cacheRead ?? 0),
+		outputTokens: usage.completion_tokens,
+		...hasExactTotal ? { totalTokens: combined } : {},
+		...cacheRead !== void 0 ? { cacheReadTokens: cacheRead } : {},
+		...reasoning !== void 0 ? { reasoningTokens: reasoning } : {}
+	};
+}
+/** Assemble the final ContentBlock for one open block. */
+function closeBlock(block) {
+	switch (block.kind) {
+		case "text": return {
+			type: "text",
+			text: block.text
+		};
+		case "reasoning": return {
+			type: "reasoning",
+			text: block.text
+		};
+		case "tool-call": return {
+			type: "tool-call",
+			id: brandString(block.callId ?? ""),
+			name: block.name ?? "",
+			arguments: block.text
+		};
+	}
+}
+/** Accept one streamed identity field for a tool call. */
+function acceptIdentity(current, incoming) {
+	return typeof incoming === "string" && incoming.length > 0 ? incoming : current;
+}
+/**
+* Consume SSE data payloads (ending with `[DONE]`) and yield StreamChunks.
+*/
+async function* translate(payloads) {
+	let nextIndex = 0;
+	let textBlock;
+	let reasoningBlock;
+	const toolBlocks = new Map();
+	let pendingFinish;
+	let pendingUsage;
+	function open(kind) {
+		const block = {
+			index: nextIndex,
+			kind,
+			text: "",
+			callId: void 0,
+			name: void 0
+		};
+		nextIndex++;
+		return block;
+	}
+	for await (const payload of payloads) {
+		if (payload === DONE) break;
+		let chunk;
+		try {
+			chunk = JSON.parse(payload);
+		} catch {
+			continue;
+		}
+		if (chunk.usage) pendingUsage = mapUsage(chunk.usage);
+		const choice = chunk.choices?.[0];
+		if (!choice) continue;
+		const delta = choice.delta;
+		if (delta) {
+			if (delta.content && delta.content.length > 0) {
+				if (textBlock === void 0) {
+					textBlock = open("text");
+					yield {
+						type: "block-start",
+						index: textBlock.index,
+						blockType: "text"
+					};
+				}
+				textBlock.text += delta.content;
+				yield {
+					type: "text-delta",
+					index: textBlock.index,
+					text: delta.content
+				};
+			}
+			if (delta.reasoning_content && delta.reasoning_content.length > 0) {
+				if (reasoningBlock === void 0) {
+					reasoningBlock = open("reasoning");
+					yield {
+						type: "block-start",
+						index: reasoningBlock.index,
+						blockType: "reasoning"
+					};
+				}
+				reasoningBlock.text += delta.reasoning_content;
+				yield {
+					type: "reasoning-delta",
+					index: reasoningBlock.index,
+					text: delta.reasoning_content
+				};
+			}
+			if (delta.tool_calls) for (const tc of delta.tool_calls) {
+				let block = toolBlocks.get(tc.index);
+				if (block === void 0) {
+					block = open("tool-call");
+					toolBlocks.set(tc.index, block);
+					block.callId = acceptIdentity(void 0, tc.id);
+					block.name = acceptIdentity(void 0, tc.function?.name);
+					yield {
+						type: "block-start",
+						index: block.index,
+						blockType: "tool-call"
+					};
+				} else {
+					block.callId = acceptIdentity(block.callId, tc.id);
+					block.name = acceptIdentity(block.name, tc.function?.name);
+				}
+				const args = tc.function?.arguments;
+				if (args) {
+					block.text += args;
+					yield {
+						type: "tool-call-delta",
+						index: block.index,
+						id: brandString(block.callId ?? ""),
+						...block.name ? { name: block.name } : {},
+						argumentsDelta: args
+					};
+				}
+			}
+		}
+		if (choice.finish_reason) pendingFinish = mapFinishReason(choice.finish_reason);
+	}
+	const allBlocks = [
+		...textBlock !== void 0 ? [textBlock] : [],
+		...reasoningBlock !== void 0 ? [reasoningBlock] : [],
+		...toolBlocks.values()
+	].sort((a, b) => a.index - b.index);
+	for (const block of allBlocks) yield {
+		type: "block-end",
+		index: block.index,
+		block: closeBlock(block)
+	};
+	if (pendingUsage !== void 0) yield {
+		type: "usage",
+		usage: pendingUsage
+	};
+	yield {
+		type: "finish",
+		reason: pendingFinish ?? { kind: "stop" }
+	};
+}
+
+//#endregion
+//#region src/adapter.ts
+/** Credential store account names for Huawei Cloud credentials. */
+const AK_ACCOUNT = "HW_ACCESS_KEY";
+const SK_ACCOUNT = "HW_SECRET_KEY";
+const TOKEN_ACCOUNT = "HW_SECURITY_TOKEN";
+/**
+* Map the gateway's `modalities` array (e.g. `["text","image"]`) to DSH's
+* `ModelModality` vocabulary. DSH currently defines only `'text' | 'image'`;
+* other modalities (audio, video) are dropped — the gateway still accepts
+* them, but the harness catalog cannot advertise them.
+*/
+function mapModalities(modalities) {
+	if (modalities === void 0 || modalities.length === 0) return {};
+	const mapped = modalities.filter((m) => m === "text" || m === "image");
+	return mapped.length > 0 ? { inputModalities: mapped } : {};
+}
+var GatewayAdapter = class GatewayAdapter extends LlmAdapter {
+	constructor(opts) {
+		super();
+		this.opts = opts;
+	}
+	providerInfo(provider) {
+		return {
+			id: provider,
+			name: "AI Shell"
+		};
+	}
+	/**
+	* Fetch the raw wire models from `/models`, cached for a short TTL so
+	* `listModels` and `resolveModel` don't each sign a separate request.
+	* Returns `[]` on any failure (missing credentials, network, non-ok) so
+	* the provider still registers — the user can type a model ID manually.
+	*/
+	async fetchWireModels() {
+		const now = Date.now();
+		if (this.modelsCache !== void 0 && this.modelsCache.expires > now) return this.modelsCache.promise;
+		const creds = this.readCredentials();
+		if (!creds?.ak) return [];
+		const promise = (async () => {
+			const response = await this.signedFetch("/models", "GET", void 0, void 0);
+			if (!response.ok) return [];
+			const json = await response.json();
+			return json.data ?? [];
+		})().catch(() => []);
+		this.modelsCache = {
+			promise,
+			expires: now + GatewayAdapter.MODELS_TTL_MS
+		};
+		return promise;
+	}
+	async listModels(provider) {
+		const wireModels = await this.fetchWireModels();
+		return wireModels.map((m) => ({
+			provider,
+			id: m.id,
+			name: m.id,
+			...mapModalities(m.modalities)
+		}));
+	}
+	async resolveModel(provider, model) {
+		const wireModels = await this.fetchWireModels();
+		const wire = wireModels.find((m) => m.id === model);
+		const base = {
+			provider,
+			id: model,
+			name: model,
+			...mapModalities(wire?.modalities)
+		};
+		if (wire === void 0) return base;
+		const contextWindow = wire.limits?.context_window;
+		const maxCompletion = wire.limits?.max_completion_tokens;
+		const efforts = wire.reasoning_effort;
+		return {
+			...base,
+			...typeof contextWindow === "number" && contextWindow > 0 ? { context: { contextWindow } } : {},
+			...typeof maxCompletion === "number" && maxCompletion > 0 ? { defaultMaxTokens: maxCompletion } : {},
+			...Array.isArray(efforts) && efforts.length > 0 ? { reasoning: { efforts: efforts.map((e) => ({
+				id: e,
+				name: e
+			})) } } : {}
+		};
+	}
+	/** Read AK/SK (and optional SecurityToken) from the credential store. */
+	readCredentials() {
+		const ak = this.opts.keyring.get(AK_ACCOUNT);
+		const sk = this.opts.keyring.get(SK_ACCOUNT);
+		if (!ak || !sk) return void 0;
+		const securityToken = this.opts.keyring.get(TOKEN_ACCOUNT);
+		return {
+			ak,
+			sk,
+			...securityToken ? { securityToken } : {}
+		};
+	}
+	/** Read credentials or throw MISSING_CREDENTIAL — every signing code path goes through here. */
+	requireCredentials() {
+		const creds = this.readCredentials();
+		if (!creds) throw new LlmError("aishell-llm: HW_ACCESS_KEY and HW_SECRET_KEY not found in credential store (service \"aishell-dsh\"); store them first", "MISSING_CREDENTIAL");
+		return creds;
+	}
+	/**
+	* Sign a request with AK/SK and fetch. Encapsulates the SDK-HMAC-SHA256
+	* signing flow: read credentials → sign → attach headers → fetch.
+	*/
+	async signedFetch(path, method, body, signal) {
+		const creds = this.requireCredentials();
+		const url = new URL(this.opts.baseUrl);
+		const headers = sign(method, url.host, path, {}, body ?? "", creds);
+		return fetch(`${this.opts.baseUrl}${path}`, {
+			method,
+			headers: {
+				host: headers.host,
+				"x-sdk-date": headers["x-sdk-date"],
+				Authorization: headers.Authorization,
+				...headers["x-security-token"] !== void 0 ? { "x-security-token": headers["x-security-token"] } : {},
+				...body !== void 0 ? { "content-type": "application/json" } : {}
+			},
+			...body !== void 0 ? { body } : {},
+			...signal ? { signal } : {}
+		});
+	}
+	async *stream(options) {
+		const body = JSON.stringify(serializeRequest(options));
+		let response;
+		try {
+			response = await this.signedFetch("/chat/completions", "POST", body, options.signal);
+		} catch (error) {
+			if (options.signal?.aborted) throw error;
+			throw new LlmError(`aishell-llm: request failed: ${error instanceof Error ? error.message : String(error)}`, "NETWORK_ERROR");
+		}
+		if (!response.ok) {
+			const text = await response.text().catch(() => "");
+			throw new LlmError(`aishell-llm: gateway returned ${response.status}: ${text}`, "PROVIDER_ERROR", { status: response.status });
+		}
+		if (!response.body) throw new LlmError("aishell-llm: gateway returned no body", "EMPTY_RESPONSE");
+		yield* translate(parseSse(response.body));
+	}
+};
+GatewayAdapter.MODELS_TTL_MS = 6e4;
+
+//#endregion
+//#region src/index.ts
+/** Default chat API base URL. */
+const DEFAULT_BASE_URL = "https://api.aishell.openbuildlab.com";
+const name = "aishell-llm";
+const inject = ["llm"];
+/** The provider route this plugin owns. */
+const PROVIDER = "aishell";
+function apply(ctx, config = {}) {
+	const store = createCredentialsStore(config.keyringService ? { service: config.keyringService } : {});
+	const adapter = new GatewayAdapter({
+		baseUrl: config.baseUrl ?? DEFAULT_BASE_URL,
+		keyring: store
+	});
+	ctx.llm.registerAdapter([PROVIDER], adapter);
+}
+
+//#endregion
+export { PROVIDER, apply, inject, name };
+//# sourceMappingURL=index.js.map

package/lib/types/adapter.d.ts

@@ -0,0 +1,48 @@
+/**
+ * aishell LLM gateway adapter for DSH.
+ *
+ * Routes model calls through the aishell gateway using direct
+ * AK/SK signing (SDK-HMAC-SHA256). No intermediate "config fetch" or API-key
+ * exchange — every request (chat + model listing) is signed with the user's
+ * AK/SK credentials read from the credential store.
+ *
+ * @module @huaweidevtools/aishell-dsh-llm/adapter
+ */
+import { LlmAdapter } from '@deepseek-ai/dsh-llm';
+import type { GenerateOptions, LlmModelInfo, LlmProviderInfo, LlmResolvedModelInfo, StreamChunk } from '@deepseek-ai/dsh-llm';
+import type { CredentialsStore } from '@huaweidevtools/aishell-dsh-credentials-store';
+/** Adapter construction options. */
+export interface GatewayAdapterOptions {
+    /** Chat API base URL (e.g. `https://api.aishell.openbuildlab.com`). */
+    readonly baseUrl: string;
+    /** Credential store for reading AK/SK. */
+    readonly keyring: CredentialsStore;
+}
+export declare class GatewayAdapter extends LlmAdapter {
+    private readonly opts;
+    /** Cached /models response (TTL-bounded so credential changes eventually refresh). */
+    private modelsCache;
+    private static readonly MODELS_TTL_MS;
+    constructor(opts: GatewayAdapterOptions);
+    providerInfo(provider: string): LlmProviderInfo;
+    /**
+     * Fetch the raw wire models from `/models`, cached for a short TTL so
+     * `listModels` and `resolveModel` don't each sign a separate request.
+     * Returns `[]` on any failure (missing credentials, network, non-ok) so
+     * the provider still registers — the user can type a model ID manually.
+     */
+    private fetchWireModels;
+    listModels(provider: string): Promise<readonly LlmModelInfo[]>;
+    resolveModel(provider: string, model: string): Promise<LlmResolvedModelInfo>;
+    /** Read AK/SK (and optional SecurityToken) from the credential store. */
+    private readCredentials;
+    /** Read credentials or throw MISSING_CREDENTIAL — every signing code path goes through here. */
+    private requireCredentials;
+    /**
+     * Sign a request with AK/SK and fetch. Encapsulates the SDK-HMAC-SHA256
+     * signing flow: read credentials → sign → attach headers → fetch.
+     */
+    private signedFetch;
+    stream(options: GenerateOptions): AsyncIterable<StreamChunk>;
+}
+//# sourceMappingURL=adapter.d.ts.map

package/lib/types/index.d.ts

@@ -0,0 +1,29 @@
+/**
+ * aishell LLM gateway adapter plugin.
+ *
+ * Registers the `aishell` provider route on `ctx.llm`, routing model calls
+ * through the aishell gateway. Every request is signed directly with the
+ * user's AK/SK (SDK-HMAC-SHA256) — no intermediate API-key exchange. AK/SK
+ * are read from the credential store (OS keyring with memory fallback) via
+ * `createCredentialsStore`.
+ *
+ * Credential management (settings namespace + watch → store transfer) lives
+ * in `aishell-credentials`; this plugin only consumes the stored credentials.
+ *
+ * @module @huaweidevtools/aishell-dsh-llm
+ */
+import type { Context } from '@deepseek-ai/cordis';
+export type { GatewayAdapter, GatewayAdapterOptions } from './adapter.ts';
+/** Plugin config. */
+export interface Config {
+    /** Chat API base URL. Default: the public aishell gateway. */
+    readonly baseUrl?: string;
+    /** Credential store service name; defaults to `'aishell-dsh'`. */
+    readonly keyringService?: string;
+}
+export declare const name = "aishell-llm";
+export declare const inject: string[];
+/** The provider route this plugin owns. */
+export declare const PROVIDER = "aishell";
+export declare function apply(ctx: Context, config?: Config): void;
+//# sourceMappingURL=index.d.ts.map

package/lib/types/serialize.d.ts

@@ -0,0 +1,13 @@
+/**
+ * Serialize harness {@link GenerateOptions} into an OpenAI-compatible wire request.
+ *
+ * Text-only: image blocks are not mapped (the gateway adapter is for text
+ * models; image support can be added later by extending `blocksToText`).
+ *
+ * @module @huaweidevtools/aishell-dsh-llm/serialize
+ */
+import type { GenerateOptions } from '@deepseek-ai/dsh-llm';
+import type { WireRequest } from './types.ts';
+/** Build the POST body for `/chat/completions`. */
+export declare function serializeRequest(options: GenerateOptions): WireRequest;
+//# sourceMappingURL=serialize.d.ts.map

package/lib/types/sse.d.ts

@@ -0,0 +1,12 @@
+/**
+ * SSE byte-stream parser for OpenAI-compatible chat-completions.
+ *
+ * Yields each event's `data` payload; the literal `[DONE]` is yielded as the
+ * final value so the caller owns final flushing. EOF before `[DONE]` raises
+ * `LlmError('STREAM_CLOSED')`.
+ *
+ * @module @huaweidevtools/aishell-dsh-llm/sse
+ */
+export declare const DONE = "[DONE]";
+export declare function parseSse(stream: ReadableStream<BufferSource>): AsyncGenerator<string>;
+//# sourceMappingURL=sse.d.ts.map

package/lib/types/translate.d.ts

@@ -0,0 +1,15 @@
+/**
+ * Translate OpenAI-compatible SSE payloads into harness {@link StreamChunk}s.
+ *
+ * One stateful block per content, reasoning, or tool-call index. Finish reason
+ * and usage are deferred until `[DONE]`, covering both finish-attached and
+ * trailing usage-only shapes.
+ *
+ * @module @huaweidevtools/aishell-dsh-llm/translate
+ */
+import type { StreamChunk } from '@deepseek-ai/dsh-llm';
+/**
+ * Consume SSE data payloads (ending with `[DONE]`) and yield StreamChunks.
+ */
+export declare function translate(payloads: AsyncIterable<string>): AsyncGenerator<StreamChunk>;
+//# sourceMappingURL=translate.d.ts.map

package/lib/types/types.d.ts

@@ -0,0 +1,128 @@
+/**
+ * OpenAI-compatible chat-completions wire format (text-only).
+ *
+ * @module @huaweidevtools/aishell-dsh-llm/types
+ */
+/** Request body for `POST {baseURL}/chat/completions`. */
+export interface WireRequest {
+    model: string;
+    messages: WireMessage[];
+    stream: true;
+    stream_options: {
+        include_usage: true;
+    };
+    tools?: WireTool[];
+    temperature?: number;
+    max_tokens?: number;
+    stop?: string[];
+}
+/** One wire message. */
+export type WireMessage = WireSystemMessage | WireUserMessage | WireAssistantMessage | WireToolMessage;
+export interface WireSystemMessage {
+    role: 'system';
+    content: string;
+}
+export interface WireUserMessage {
+    role: 'user';
+    content: string;
+}
+export interface WireAssistantMessage {
+    role: 'assistant';
+    content?: string | null;
+    tool_calls?: WireAssistantToolCall[];
+}
+export interface WireToolMessage {
+    role: 'tool';
+    tool_call_id: string;
+    content: string;
+}
+export interface WireAssistantToolCall {
+    id: string;
+    type: 'function';
+    function: {
+        name: string;
+        arguments: string;
+    };
+}
+export interface WireTool {
+    type: 'function';
+    function: {
+        name: string;
+        description: string;
+        parameters: Record<string, unknown>;
+    };
+}
+/** One SSE chunk from the chat-completions stream. */
+export interface WireChunk {
+    choices?: WireChoice[];
+    usage?: WireUsage;
+}
+/** One model entry from the GET /models response (OpenAI-compatible). */
+export interface WireModel {
+    id: string;
+    object?: string;
+    owned_by?: string;
+    created?: number;
+    capabilities?: WireModelCapabilities;
+    modalities?: string[];
+    reasoning_effort?: string[];
+    limits?: WireModelLimits;
+    pricing?: WireModelPricing;
+}
+export interface WireModelCapabilities {
+    text?: boolean;
+    image?: boolean;
+    video?: boolean;
+    audio?: boolean;
+    streaming?: boolean;
+    thinking?: boolean;
+    tool_call?: boolean;
+    embedding?: boolean;
+}
+export interface WireModelLimits {
+    context_window?: number;
+    max_prompt_tokens?: number;
+    max_completion_tokens?: number;
+}
+export interface WireModelPricing {
+    currency?: string;
+    unit?: string;
+    input?: number;
+    cached_input?: number;
+    output?: number;
+}
+/** Response body from GET /models. */
+export interface WireModelList {
+    data: WireModel[];
+}
+export interface WireChoice {
+    delta?: WireDelta;
+    finish_reason?: string | null;
+}
+export interface WireDelta {
+    role?: string;
+    content?: string;
+    reasoning_content?: string;
+    tool_calls?: WireDeltaToolCall[];
+}
+export interface WireDeltaToolCall {
+    index: number;
+    id?: string;
+    type?: 'function';
+    function?: {
+        name?: string;
+        arguments?: string;
+    };
+}
+export interface WireUsage {
+    prompt_tokens: number;
+    completion_tokens: number;
+    total_tokens?: number;
+    prompt_tokens_details?: {
+        cached_tokens?: number;
+    };
+    completion_tokens_details?: {
+        reasoning_tokens?: number;
+    };
+}
+//# sourceMappingURL=types.d.ts.map

package/package.json

@@ -1,6 +1,50 @@
 {
   "name": "@huaweidevtools/aishell-dsh-llm",
-  "version": "0.0.0-stage",
-  "stub": true,
-  "description": "Temporary package placeholder for staged publishing"
+  "description": "aishell LLM gateway adapter for DSH — OpenAI-compatible, Huawei Cloud AK/SK auth via credential store",
+  "version": "0.0.1-beta.0",
+  "publishConfig": {
+    "access": "public"
+  },
+  "repository": {
+    "type": "git",
+    "url": "git+https://github.com/yusheng-g/aishell-dsh-plugins.git",
+    "directory": "packages/aishell-llm"
+  },
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./lib/types/index.d.ts",
+      "default": "./lib/index.js"
+    }
+  },
+  "files": [
+    "lib/index.js",
+    "lib/types/**/*.d.ts",
+    "cordis.patch.yml"
+  ],
+  "license": "Apache-2.0",
+  "dependencies": {
+    "eventsource-parser": "^3.0.0",
+    "@huaweidevtools/aishell-dsh-credentials-store": "0.0.1-beta.0"
+  },
+  "peerDependencies": {
+    "@deepseek-ai/cordis": "^4.0.4",
+    "@deepseek-ai/dsh-brand": "0.2.0-rc.1 || 0.2.1-alpha.1 || ^0.2.1",
+    "@deepseek-ai/dsh-llm": "0.2.0-rc.1 || 0.2.1-alpha.1 || ^0.2.1"
+  },
+  "devDependencies": {
+    "@deepseek-ai/cordis": "^4.0.4",
+    "@deepseek-ai/dsh-brand": "0.2.0-rc.1 || 0.2.1-alpha.1 || ^0.2.1",
+    "@deepseek-ai/dsh-llm": "0.2.0-rc.1 || 0.2.1-alpha.1 || ^0.2.1"
+  },
+  "dsh": {
+    "bundle": {
+      "patch": "./cordis.patch.yml"
+    }
+  },
+  "scripts": {
+    "build:bundle": "tsdown"
+  }
 }