@deepseek-ai/dsh-tool-skill 0.1.5-rc.1 → 0.1.6-alpha.1

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (4) hide show
  1. package/README.i18n.yaml +2 -2
  2. package/README.md +2 -2
  3. package/README.zh.md +5 -5
  4. package/package.json +14 -14
package/README.i18n.yaml CHANGED
@@ -2,5 +2,5 @@
2
2
  # side as of the last confirmed-consistent state. Both languages carry equal authority;
3
3
  # after editing either side, bring the other along and re-record with:
4
4
  # pnpm run verify-translation-pairing --write packages/skill/tool-skill/README.md
5
- README.md: 090dd3474087a9fb0c944bb88081c759cbfed20c
6
- README.zh.md: 7dcf1c216e12391e1cc6810579fbc76fcaba7890
5
+ README.md: eeee637c25bc38138deaf907e6bf73202f408b96
6
+ README.zh.md: f2c1adf49a838e84fd488f0cb803d3aa30f8b8ca
package/README.md CHANGED
@@ -9,7 +9,7 @@ English | [中文](README.zh.md)
9
9
 
10
10
  ## Summary
11
11
 
12
- Agents can discover and load skills during a session. Before the first request, they receive a durable catalog of available skill names and capped descriptions, and can use the `skill` tool to load full instructions. Users can invoke a skill with `/name`, which injects the same instructions into that step. Catalog changes append a complete replacement, including an empty catalog that retires old names; configure `catalogDescriptionMaxLength` to limit each description.
12
+ Agents can discover and load skills during a session. Before the first request, when model-invocable skills exist and the `skill` tool is visible, they receive a durable catalog of available skill names and capped descriptions, and can use the `skill` tool to load full instructions. Users can invoke a user-invocable skill with `/name`, which injects the same instructions into that step. Catalog changes append a complete replacement, including an empty catalog that retires old names; configure `catalogDescriptionMaxLength` to limit each description.
13
13
 
14
14
  ## Table of Contents
15
15
 
@@ -56,7 +56,7 @@ The generated [configuration catalog](../../../docs/config-catalog.md#deepseek-a
56
56
 
57
57
  ### Observable success and failures
58
58
 
59
- Loading a listed skill returns its full instructions; the model sees one canonical shape whether the load came from the tool or from a user's explicit invocation. An invalid name reports `Error: invalid skill name "<name>"`, an unknown name reports the skill is unknown or no longer available, and a skill disabled for model invocation reports it is not available for model invocation. The catalog is omitted entirely only when no model-invocable skills exist and none was ever published; a later visibility loss — the `skill` tool hidden or shadowed by a same-name scoped tool — instead appends an empty retirement catalog, as when every skill is removed.
59
+ Loading a listed skill returns its full instructions; the model sees one canonical shape whether the load came from the tool or from a user's explicit invocation. An invalid name reports `Error: invalid skill name "<name>"`, an unknown name reports the skill is unknown or no longer available, and a skill disabled for model invocation reports it is not available for model invocation. The catalog is omitted entirely when no catalog was ever published and either no model-invocable skills exist or the `skill` tool is hidden or shadowed; after a catalog has been published, either visibility loss — the `skill` tool hidden or shadowed by a same-name scoped tool — or removal of every skill instead appends an empty catalog that retires older names.
60
60
 
61
61
  -----
62
62
 
package/README.zh.md CHANGED
@@ -1,5 +1,5 @@
1
1
  ---
2
- description: "面向模型的 skill 目录与加载工具，供了解 agent 看到什么、或配置会话 skill 目录的用户与维护者阅读。"
2
+ description: "面向模型的 skill（技能）目录与加载工具，供希望了解 agent（智能体）看到的内容或配置会话 skill 目录的用户与维护者阅读。"
3
3
  kind: "package-reference"
4
4
  ---
5
5
 
@@ -9,7 +9,7 @@ kind: "package-reference"
9
9
 
10
10
  ## 概述
11
11
 
12
- agent（智能体）可以在会话期间发现并加载 skill（技能）。在首次请求前，它们会收到一份持久目录，列出可用 skill 的名称与有长度上限的描述，并可用 `skill` 工具加载完整指令。用户可以用 `/name` 调用某个 skill，把相同的指令注入该步骤。目录变更会追加一份完整替换，其中空目录会停用旧名称；可配置 `catalogDescriptionMaxLength` 来限制每条描述的长度。
12
+ agent 可以在会话期间发现并加载 skill。在首次请求前，如果存在模型可调用 skill 且 `skill` 工具可见，agent 会收到一份持久目录，列出可用 skill 的名称与有长度上限的描述，并可用 `skill` 工具加载完整指令。用户可以用 `/name` 调用某个用户可调用的 skill，把相同的指令注入该步骤。目录变更会追加一份完整替换，其中空目录会停用旧名称；可配置 `catalogDescriptionMaxLength` 来限制每条描述的长度。
13
13
 
14
14
  ## 目录
15
15
 
@@ -51,12 +51,12 @@ agent（智能体）可以在会话期间发现并加载 skill（技能）。在
51
51
 
52
52
  - **会话目录。** 当存在模型可调用 skill 且 `skill` 工具可见时，agent 会在首次请求前收到一条持久的用户角色消息，列出每个 skill 的名称与有长度上限的描述；该消息告诉模型在着手任务前先用工具加载 skill，且绝不能仅凭摘要推断指令。
53
53
  - **加载工具。** 模型以精确的 skill 名称调用 `skill`，并收到完整指令正文以及规范的 `<skill_content>` 块中的资源指引；该结果作为普通工具历史保留。
54
- - **用户显式调用。** 直接用户输入中的 `/name` token 若指名某个用户可调用 skill，会把该 skill 的指令注入当轮次，而无需模型自行加载。
54
+ - **用户显式调用。** 直接用户输入中的 `/name` token 若指名某个用户可调用 skill，会把该 skill 的指令注入该步骤，而无需模型自行加载。
55
55
  - **实时目录更新。** 后续成员关系、描述或可见性变化会追加完整的替换目录；删除全部 skill 时会追加空目录，停用较早的名称。
56
56
 
57
57
  ### 可观察的成功与失败
58
58
 
59
- 加载列出的 skill 会返回其完整指令；无论加载来自工具还是用户的显式调用，模型看到的都是同一种规范形态。无效名称会报告 `Error: invalid skill name "<name>"`，未知名称会报告该 skill 未知或已不可用，被禁用模型调用的 skill 会报告其不可用于模型调用。只有不存在模型可调用 skill 且从未发布过目录时，目录才会被整体省略；此后的可见性丧失——`skill` 工具被隐藏或被同名作用域工具遮蔽——会改为追加空目录来停用旧名称，与删除全部 skill 时相同。
59
+ 加载列出的 skill 会返回其完整指令；无论加载来自工具还是用户的显式调用，模型看到的都是同一种规范形态。无效名称会报告 `Error: invalid skill name "<name>"`，未知名称会报告该 skill 未知或已不可用，被禁用模型调用的 skill 会报告其不可用于模型调用。如果从未发布过目录，并且不存在模型可调用 skill，或 `skill` 工具被隐藏或遮蔽，则会整体省略目录；目录发布后，无论可见性丧失——`skill` 工具被隐藏或被同名作用域工具遮蔽——还是删除全部 skill，都会改为追加空目录来停用旧名称。
60
60
 
61
61
  -----
62
62
 
@@ -77,7 +77,7 @@ agent（智能体）可以在会话期间发现并加载 skill（技能）。在
77
77
  | 文件 | 职责 |
78
78
  |---|---|
79
79
  | [`src/index.ts`](src/index.ts) | 插件入口：工具注册、目录与手势 pre-step 监听器、渲染与 digest |
80
- | — | 不发布运行时不变式伴生入口；这个模型侧 adapter 没有独立 lifecycle stream；执行关系由它调用的 capability seam 负责。 |
80
+ | — | 不发布运行时不变式伴生入口；这个面向模型的适配器没有独立的生命周期流；执行关系由它调用的能力 seam 负责。 |
81
81
 
82
82
  ### 目录生命周期
83
83
 
package/package.json CHANGED
@@ -1,7 +1,7 @@
1
1
  {
2
2
  "name": "@deepseek-ai/dsh-tool-skill",
3
3
  "description": "Model-facing skill loading tool for the DeepSeek Harness",
4
- "version": "0.1.5-rc.1",
4
+ "version": "0.1.6-alpha.1",
5
5
  "publishConfig": {
6
6
  "access": "public"
7
7
  },
@@ -27,24 +27,24 @@
27
27
  ],
28
28
  "license": "MIT",
29
29
  "peerDependencies": {
30
- "@deepseek-ai/dsh-agent": "^0.1.5-rc.1",
31
- "@deepseek-ai/dsh-llm": "^0.1.5-rc.1",
32
- "@deepseek-ai/dsh-skill": "^0.1.5-rc.1",
33
- "@deepseek-ai/cordis": "^4.0.2",
34
- "@deepseek-ai/dsh-tools": "^0.1.5-rc.1"
30
+ "@deepseek-ai/dsh-agent": "^0.1.6-alpha.1",
31
+ "@deepseek-ai/dsh-llm": "^0.1.6-alpha.1",
32
+ "@deepseek-ai/dsh-skill": "^0.1.6-alpha.1",
33
+ "@deepseek-ai/dsh-tools": "^0.1.6-alpha.1",
34
+ "@deepseek-ai/cordis": "^4.0.2"
35
35
  },
36
36
  "dependencies": {
37
37
  "@deepseek-ai/schemastery": "^3.18.2"
38
38
  },
39
39
  "devDependencies": {
40
- "@deepseek-ai/dsh-agent": "^0.1.5-rc.1",
41
- "@deepseek-ai/dsh-agent-loop-testkit": "^0.1.5-rc.1",
42
- "@deepseek-ai/dsh-llm": "^0.1.5-rc.1",
43
- "@deepseek-ai/dsh-scope": "^0.1.5-rc.1",
44
- "@deepseek-ai/dsh-skill-filesystem": "^0.1.5-rc.1",
45
- "@deepseek-ai/dsh-skill": "^0.1.5-rc.1",
46
- "@deepseek-ai/dsh-session": "^0.1.5-rc.1",
47
- "@deepseek-ai/dsh-tools": "^0.1.5-rc.1",
40
+ "@deepseek-ai/dsh-agent": "^0.1.6-alpha.1",
41
+ "@deepseek-ai/dsh-agent-loop-testkit": "^0.1.6-alpha.1",
42
+ "@deepseek-ai/dsh-llm": "^0.1.6-alpha.1",
43
+ "@deepseek-ai/dsh-scope": "^0.1.6-alpha.1",
44
+ "@deepseek-ai/dsh-session": "^0.1.6-alpha.1",
45
+ "@deepseek-ai/dsh-skill": "^0.1.6-alpha.1",
46
+ "@deepseek-ai/dsh-skill-filesystem": "^0.1.6-alpha.1",
47
+ "@deepseek-ai/dsh-tools": "^0.1.6-alpha.1",
48
48
  "@deepseek-ai/cordis": "^4.0.2"
49
49
  }
50
50
  }