chore(release): prepare v1.7.2

This commit is contained in:
Ma
2026-07-28 17:18:39 +08:00
parent 6e4ce0050e
commit c7851b94ad
8 changed files with 54 additions and 14 deletions
+20
View File
@@ -2,6 +2,26 @@
[中文](CHANGELOG.md) | English
## v1.7.2
### Release Focus
Agent Skills compatibility and a single Skill runtime. InkOS now consumes standard `SKILL.md` packages directly. The Chat Agent can activate a Skill from user intent, while users can force one from the composer `+` button or with `@skill-id`. The former InkOS-private Skill protocol, keyword triggers, and parallel context planner have been removed instead of being layered beside the new system.
### Agent Skills
- Added the typed `use_skill` tool to the pi-agent loop: the model sees available Skill names and descriptions, then activates professional guidance through a structured tool call
- Studio can import a complete folder containing `SKILL.md`; static references are preserved under the project's `.agents/skills/`, while external scripts are never executed automatically
- Added standard project discovery from `skills/` and `.agents/skills/`, plus user discovery from `~/.agents/skills/` and `~/.openclaw/skills/`
- Users can select and force a Skill in the Chat composer; one-turn intent activation expires instead of leaking into later turns or another book
- Project Settings shows each Skill's source and can delete project imports; deletion immediately removes the Skill from both the Agent catalog and composer
### Compatibility
- Removed the legacy `.inkos/skills` loader and InkOS-only fields such as `whenToUse`, `promptPacks`, `toolHints`, and `contextNeeds`
- Prompt packs remain a separate, Studio-editable subsystem under Project Settings instead of masquerading as Skills
- Migrate old Skills to a standard folder containing `SKILL.md` and static references, then import it in Studio or place it under `.agents/skills/`
## v1.7.1
### Release Focus
+20
View File
@@ -2,6 +2,26 @@
[English](CHANGELOG.en.md) | 中文
## v1.7.2
### Release Focus
Agent Skills 兼容与单一 Skill 内核更新:InkOS 现在直接读取标准 `SKILL.md` 能力包,Chat Agent 可以根据用户意图自主调用,也可以由用户通过输入框 `+` 按钮或 `@skill-id` 强制启用。旧的 InkOS 私有 Skill 协议、关键词触发器和并行上下文规划器已被完整替换,不再叠加两套实现。
### Agent Skills
- 新增基于 pi-agent 工具循环的 `use_skill`:模型先读取可用 Skill 的名称与描述,再通过结构化工具调用激活所需专业能力
- Studio 支持导入包含 `SKILL.md` 的完整文件夹,静态参考资料会一并保存到项目 `.agents/skills/`;外部脚本不会被自动执行
- 支持项目 `skills/``.agents/skills/`,以及用户级 `~/.agents/skills/``~/.openclaw/skills/` 等标准位置
- 用户可在 Chat 输入区选择并强制启用 Skill;一次性意图激活不会永久污染后续会话或切换书籍后的上下文
- Studio 项目设置统一展示 Skill 来源,并支持删除项目导入的 Skill;删除后立即从 Agent 与选择列表中移除
### Compatibility
- 删除旧 `.inkos/skills` 私有加载路径,以及 `whenToUse``promptPacks``toolHints``contextNeeds` 等 InkOS 专用字段
- 提示词包继续由 Studio“项目设置 → 提示词”独立管理,不再伪装成 Skill 能力
- 旧 Skill 请迁移为标准目录:保留 `SKILL.md` 与静态参考资料,通过 Studio 重新导入或放入 `.agents/skills/`
## v1.7.1
### Release Focus
+1 -1
View File
@@ -1,6 +1,6 @@
{
"name": "inkos",
"version": "1.7.1",
"version": "1.7.2",
"private": true,
"description": "Story Creation AI Agent for long-form and short fiction, scripts, interactive games, and IP content",
"keywords": [
+3 -3
View File
@@ -1,6 +1,6 @@
{
"name": "@actalk/inkos",
"version": "1.7.1",
"version": "1.7.2",
"description": "Story Creation AI Agent CLI for long-form fiction, InkOS Short, fan works, continuations, covers, and interactive worlds.",
"keywords": [
"story-creation-ai-agent",
@@ -49,8 +49,8 @@
"typecheck": "tsc --noEmit"
},
"dependencies": {
"@actalk/inkos-core": "1.7.1",
"@actalk/inkos-studio": "1.7.1",
"@actalk/inkos-core": "1.7.2",
"@actalk/inkos-studio": "1.7.2",
"commander": "^13.0.0",
"dotenv": "^16.4.0",
"epub-gen-memory": "^1.0.10",
+1 -1
View File
@@ -1,6 +1,6 @@
{
"name": "@actalk/inkos-core",
"version": "1.7.1",
"version": "1.7.2",
"description": "InkOS core engine for story creation agents: governed context, writing, audit, revision, covers, and interactive worlds",
"keywords": [
"story-creation-ai-agent",
+2 -2
View File
@@ -1,6 +1,6 @@
{
"name": "@actalk/inkos-studio",
"version": "1.7.1",
"version": "1.7.2",
"description": "InkOS Studio — web workbench for story creation, InkOS Short, covers, and open-world interactive play",
"type": "module",
"main": "dist/api/index.js",
@@ -31,7 +31,7 @@
"typecheck": "tsc --noEmit && tsc -p tsconfig.server.json --noEmit"
},
"dependencies": {
"@actalk/inkos-core": "1.7.1",
"@actalk/inkos-core": "1.7.2",
"@base-ui/react": "^1.3.0",
"@fontsource-variable/geist": "^5.2.8",
"@hono/node-server": "^1.13.0",
+3 -3
View File
@@ -15,10 +15,10 @@ importers:
packages/cli:
dependencies:
'@actalk/inkos-core':
specifier: 1.7.1
specifier: 1.7.2
version: link:../core
'@actalk/inkos-studio':
specifier: 1.7.1
specifier: 1.7.2
version: link:../studio
commander:
specifier: ^13.0.0
@@ -110,7 +110,7 @@ importers:
packages/studio:
dependencies:
'@actalk/inkos-core':
specifier: 1.7.1
specifier: 1.7.2
version: link:../core
'@base-ui/react':
specifier: ^1.3.0
+4 -4
View File
@@ -1,7 +1,7 @@
---
name: inkos
description: Story Creation and Translation AI Agent with Studio Chat, CLI, and TUI - use for long-form novels, short fiction, scripts, storyboards, interactive-film projects, open-world / branching play, fan fiction, spinoffs, style imitation, continuations, covers, and multilingual EPUB/PDF/TXT/Markdown translation. Includes Agent Skills, traceable research, governed context, persistent story state, multi-model routing, image services, and InkOS Studio.
version: 2.7.0
version: 2.8.0
metadata: { "openclaw": { "emoji": "📖", "requires": { "bins": ["inkos", "node"], "env": ["OPENAI_API_KEY"] }, "primaryEnv": "OPENAI_API_KEY", "homepage": "https://github.com/Narcooo/inkos", "install": [{ "id": "npm", "kind": "node", "package": "@actalk/inkos", "label": "Install InkOS (npm)" }] } }
---
@@ -16,9 +16,9 @@ Long-form writing still uses the chapter pipeline internally:
Truth files are persisted as schema-validated JSON (`story/state/*.json`) with markdown projections for human readability. SQLite temporal memory database (`story/memory.db`) enables relevance-based retrieval on Node 22+.
## v1.7.1 Mental Model
## v1.7.2 Mental Model
Treat InkOS as a confirmable action system, not a bag of prompt shortcuts. v1.7.1 adds non-canonical narrative forecasting, background production tasks that do not block discussion, retryable task state, whole-book backup / restore, and rollback-safe latest-chapter deletion.
Treat InkOS as a confirmable action system, not a bag of prompt shortcuts. v1.7.2 replaces the former InkOS-private Skill protocol with standard AgentSkills / OpenClaw `SKILL.md` packages. Skills provide professional guidance and static references only: they can be selected explicitly or activated by the Chat Agent through `use_skill`, but they never grant execution permissions.
- Natural-language requests should go through Studio Chat / TUI / `inkos interact` whenever possible.
- Do not infer success from assistant prose. A book, short, cover, or play step is complete only when the corresponding tool result and files exist.
@@ -38,7 +38,7 @@ Treat InkOS as a confirmable action system, not a bag of prompt shortcuts. v1.7.
- Prompt packs are user-tunable in Studio Project Settings. Project overrides are saved under `prompt/<pack>/<prompt>.md`; do not edit generated artifacts just to change system behavior.
- Long-form chapter revision from Chat passes the current user instruction into the reviser as a one-off brief. If the revision is not applied, inspect the returned gate metrics and remaining audit issues before claiming it was fixed.
v1.7.0 keeps the v1.6 interactive-film and Agent Skills model, then adds complete multilingual translation/localization, English-native short/script/storyboard/interactive-film paths, existing-novel import from Chat, configurable review and revision gates, abortable long tasks, recoverable write locks, material archive/retrieval, and Studio-editable prompt packs. Still surface unresolved review or execution issues plainly instead of claiming they were fixed.
v1.7.0 added complete multilingual translation/localization, English-native short/script/storyboard/interactive-film paths, existing-novel import from Chat, configurable review and revision gates, abortable long tasks, recoverable write locks, material archive/retrieval, and Studio-editable prompt packs. v1.7.1 added non-canonical narrative forecasting, background production tasks, retryable task state, whole-book backup / restore, and rollback-safe latest-chapter deletion. Still surface unresolved review or execution issues plainly instead of claiming they were fixed.
## When to Use InkOS