Commit Graph

169 Commits

Author SHA1 Message Date
fe6fba3958 feat(multi_agent): 提取prompt、恢复任务、优化侧边栏与sleep工具
- 将多智能体/子智能体prompt从代码提取到prompts/multi_agent/和prompts/sub_agent/

- 多智能体模式添加可用的子智能体动态prompt并冻结

- 重启后自动从conversation.json恢复多智能体idle任务

- 修复新对话侧边栏显示其他对话子智能体的问题

- 简化子智能体弹窗输出样式

- sleep工具在多智能体模式下移除wait_sub_agent_ids参数
2026-07-13 14:01:48 +08:00
692748d567 fix(multi-agent): 修复消息渲染、分类、存储与滚动抖动
- 子智能体消息渲染为左侧对话气泡
- 多智能体通知细分为进度/完成/提问三类
- 修复多智能体会话存储到 mutiagents 目录并迁移数据
- 修复 token-statistics API 支持多智能体会话
- 修复 ma_debug 未定义报错
- 移除 .message-block content-visibility 修复滚动抖动
2026-07-12 23:09:56 +08:00
7f2ad9144d fix(multi-agent): 修复子智能体状态误判、后台任务隔离与 idle 等待异常 2026-07-12 18:39:21 +08:00
f652118527 fix(multi-agent): 跨事件循环问答、线程安全取消、任务结束汇报格式、工具参数可选 2026-07-12 15:09:16 +08:00
811974d6e7 feat(multi-agent): 在现有架构上直接实现多智能体模式实验功能
放弃完全隔离策略,改为在现有 MainTerminal/SubAgentManager/SubAgentTask 主链路
按对话级开关 metadata.multi_agent_mode=true 增加多智能体分支。

新增模块:
- modules/multi_agent/__init__.py: 模块入口
- modules/multi_agent/role_store.py: 角色 Markdown Frontmatter 解析与归档
- modules/multi_agent/state.py: 多智能体会话状态机与消息格式化
- modules/multi_agent/prompts.py: 主智能体(Team Leader) + 子智能体提示词
- modules/multi_agent/tools.py: 9 个主智能体工具 + 4 个子智能体工具定义
- server/multi_agent.py: /multiagent/new 页面 + /api/multiagent/* 蓝图

现有代码改动:
- modules/sub_agent/task.py: 扩展 multi_agent_mode/multi_agent_state/display_name 字段,
  增加 ask_master/ask_other_agent/answer_other_agent/list_active_sub_agents 工具处理逻辑,
  子智能体自然结束 assistant 输出即本轮结束(不调用 finish_task),上下文保留。
- modules/sub_agent/manager.py: create_sub_agent 增加 multi_agent_mode/role_id/display_name 参数,
  增加 get_or_create_multi_agent_state/get_multi_agent_state/inject_message_to_sub_agent/_on_multi_agent_task_done 方法。
- core/main_terminal_parts/tools_definition/agent_tools.py: 多智能体模式下用 modules.multi_agent.tools 替换旧版工具集。
- core/main_terminal_parts/context/messages.py: 多智能体模式下追加 Team Leader 系统提示词。
- core/main_terminal_parts/tools_execution.py: create_sub_agent handler 增加多智能体分支,新增 send_message_to_sub_agent/ask_sub_agent/answer_sub_agent_question/create_custom_agent/list_agents/list_active_sub_agents handler。
- core/web_terminal.py: load_conversation 时检测 metadata.multi_agent_mode 设置 self.multi_agent_mode。
- server/app_legacy.py: 注册 multi_agent_bp 蓝图。

前端改动:
- static/src/auth/LoginApp.vue: 登录页增加'多智能体模式(beta)'按钮
- static/src/app/methods/ui/route.ts: 识别 /multiagent/new 和 /multiagent/conv_xxx 路径,进入多智能体模式并创建带 metadata.multi_agent_mode=true 的对话
- static/src/app/state.ts: 增加 multiAgentMode 状态字段

数据:
- ~/.astrion/astrion/host/mutiagents/agents/: 4 个预置角色 ui-operator / full-stack-engineer / code-reviewer / researcher
- ~/.astrion/astrion/host/mutiagents/conversations/: 会话数据

验证:所有 Python 文件语法检查通过;冒烟测试 test.test_server_refactor_smoke 6 项全通过;前端构建通过(6.04s);模块导入与功能断言测试全部通过。
2026-07-12 03:26:02 +08:00
cc8d5817aa feat(ui): 极简模式展开高度限制开关与个人空间持久化 2026-07-11 19:58:35 +08:00
af56de6812 fix(sub_agent): 修复 wait_for_completion 因 _load_state 替换 tasks 字典导致的引用失效 2026-07-11 19:19:41 +08:00
6618ab03dc feat(personalization): 将堆叠块显示模式保存到后端 personalization.json 2026-07-11 11:14:48 +08:00
5244768494 feat(sidebar): 侧边栏按工作区/项目分组显示对话
- 新增个人化设置 group_sidebar_by_workspace / sidebar_pinned_workspaces /
  sidebar_workspace_order,支持分组视图、置顶、排序
- 后端 conversation API 支持 workspace_id 参数,用于获取/创建/加载/删除
  指定工作区的对话,避免切换视图时 session 漂移
- server/context.py 新增 update_session 参数,临时解析目标 terminal 时不污染
  当前会话
- 前端 ConversationSidebar 实现分组折叠、运行中指示器、新建/更多操作菜单、
  动画同步
- config/terminal.py 默认 TERMINAL_SANDBOX_MODE 由 docker 改为 host
2026-07-11 01:51:01 +08:00
19746e67d3 feat(personalization): 将回答时必须考虑的信息从条列式改为大段文字输入 2026-07-10 15:40:18 +08:00
e3c897947f refactor(paths): 统一迁移运行态数据路径到 Astrion 命名空间
- 工作区内部路径 .agents/ -> .astrion/
- 默认运行态数据根 ~/.agents/agents/ -> ~/.astrion/astrion/
- 环境变量 AGENTS_DATA_ROOT -> ASTRION_DATA_ROOT(无兼容)
- 同步更新代码、测试、文档与脚本中的路径引用
- 新增 .gitignore 忽略 .astrion/ 与 .agents/ 运行态目录
2026-07-10 00:35:15 +08:00
6a2cb820bc fix(versioning): 修复浅备份文件修改归属错误及列表/详情不一致 2026-07-09 18:45:40 +08:00
765196ea3c feat(skills): host 模式 skill 归档目录迁移到运行态 agentskills 2026-07-09 18:42:42 +08:00
3db4e0beb4 feat(sandbox): macOS host sandbox deny-list for sensitive paths
- Add default deny_read_paths/deny_read_regexes to host_sandbox_policy.json
- Generate deny rules for ~/.ssh, ~/.aws, ~/.kube, keychains, .env, etc.
- Readonly sandbox: global read + deny sensitive + write only /dev/null
- Workspace-write sandbox: global read + deny sensitive + writable paths
- Keep workspace files readable by re-allowing workspace subpath after deny
- Backend API validates path authorization against deny list
- Path auth dialog hint explains readable paths behavior
2026-07-09 16:25:04 +08:00
478d24c4dd feat(versioning): 重构版本控制为对话级浅备份 + 可选完全备份
本次变更是对版本控制功能的大范围重构,核心目标:
- 默认使用浅备份(只追踪 write_file/edit_file 实际修改的文件),避免大工作区首次初始化耗时十几秒;
- 保留可选的完全备份(git write-tree),不污染 git log;
- 回溯时统一选择范围(仅对话 / 对话+工作区)和模式(覆盖 / 复制);
- 修复复制对话、侧边栏同步、新建对话延迟、diff 统计与渲染等连锁问题。

后端改动:
- modules/shallow_versioning.py(新增)
  - 实现 ShallowVersioningManager:track_edit、make_snapshot、rewind、list_snapshots。
  - 基于 difflib.SequenceMatcher 计算真实 insertions/deletions,replace 场景正确显示 +N/-M。
  - diff 详情使用结构化 add/remove/context 行,不再解析 unified_diff 文本。
- modules/versioning_manager.py
  - 完全备份从 git commit 改为 git write-tree,字段 commit/parent_commit 改为 tree_hash/parent_tree_hash。
  - 恢复改用 read-tree + checkout-index;create_checkpoint 支持 diff_summary 参数。
  - get_checkpoint_detail 保留 API 字段名 last_commit/restored_commit 以兼容前端。
- core/main_terminal_parts/tools_execution.py
  - write_file/edit_file 在执行前调用浅备份 track_edit,保留编辑前状态。
- server/chat_flow_task_main.py
  - 消息轮次结束时根据 backup_mode 选择完全备份 create_checkpoint 或浅备份 make_snapshot。
  - 浅备份模式下同时写入 ConversationVersioningManager 检查点行,让前端能统一展示。
- server/conversation.py
  - restore/detail/list API 支持 shallow 模式;list/detail 对浅备份实时重新计算统计。
  - 创建对话时读取 personalization 的 versioning_backup_mode,shallow 模式固定 tracking_mode 为 conversation_only。
- core/web_terminal.py
  - _ensure_conversation_versioning_enabled 读取默认备份模式并写入 versioning meta。
- modules/file_manager/crud_mixin.py / replace_mixin.py
  - write_file / edit_file 返回 original_file / new_file / replacement_details(真实旧行/新行)。
- modules/personalization_manager.py
  - 新增 versioning_enabled_by_default、versioning_backup_mode(shallow/full)。
- utils/perf_log.py(新增)
  - 创建新对话全链路性能日志,用于定位延迟瓶颈。
- utils/conversation_manager/{index,list_search,crud}_mixin.py
  - 加性能日志;index_mixin 主动检查新增对话文件,修复侧边栏不同步。

前端改动:
- static/src/components/overlay/VersioningDialog.vue
  - 去掉顶部管理范围下拉;底部新增回溯范围与回溯模式选择。
  - diff 详情改为 div.diff-line 网格布局(marker + content),支持横向滚动,状态显示缩写 A/M/D。
  - 修复 pre 标签空白黑行、容器被挤窄导致折行等问题。
- static/src/app/methods/versioning.ts
  - 移除 mismatch 弹窗逻辑;支持 copy 模式;selectVersioningCheckpoint 增强错误捕获。
- static/src/components/personalization/PersonalizationDrawer.vue
  - 「工作区与权限」新增版本控制默认开关与备份方式选择。
- static/src/stores/personalization.ts
  - 新增 versioning_enabled_by_default、versioning_backup_mode 字段。
- static/src/components/chat/actions/ToolAction.vue / toolRenderers.ts
  - edit_file / write_file 工具展示基于 result.details / result.new_file 的真实替换结果。
- static/src/app/state.ts / watchers.ts / App.vue
  - 新增 versioningInitializingBackupToastId,完全备份初始化时显示 toast。
- static/src/app/methods/conversation/action.ts / message/send.ts
  - host + 完全备份 + 默认开启时,创建新对话/从 /new 发送首条消息显示「正在初始化备份」toast。

验证:
- /Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/bin/python3.9 -m unittest test.test_server_refactor_smoke 通过。
- npm run build 通过。
- 宿主机与 docker 模式新建对话、编辑文件、回溯、复制对话均验证通过。
2026-07-07 12:07:44 +08:00
ae5676cd30 feat(ui): add assistant status avatar 2026-07-04 22:33:00 +08:00
38ff51416f fix(mcp_client_manager): restore missing imports and dataclass decorators after split 2026-06-25 11:56:19 +08:00
72a49a7c8d fix: 修复对话计时器持久化、子智能体工具美化与状态查询,统一后台通知池
- 新增 server/work_timer.py,在对话真正空闲(无前台任务、无后台子智能体/后台命令/压缩)时持久化 work_timer,并同步内存副本,解决刷新后计时器回退问题。
- server/chat_flow_task_main.py / chat_flow.py / tasks/models.py 在任务结束/取消时按空闲判定决定是否持久化。
- 前端 lifecycle.ts 在仍有后台任务时不提前停止计时器。
- 子智能体工具(create/terminate/get_status)渲染美化,task 参数放在顶部元信息区。
- 子智能体状态查询支持返回「已完成」「已终止」「不存在」;修复 wait_for_completion 在 final_result 就绪前返回导致的「ID 被占用」误报。
- 统一后台完成通知池 poll_completion_notifications,合并子智能体与后台 run_command 两路轮询,避免逐条触发工作循环与单工作区互斥冲突。
- 删除本次新增的各类 debug_log / notify_pool_log 调用及辅助脚本。
2026-06-25 04:16:44 +08:00
d1a9a51bb8 fix(terminal_ops): ssh 远程命令中避免重写 /Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/bin/python3.9/pip 路径 2026-06-24 17:33:12 +08:00
e90610ed05 fix(frontend): 修复刷新后AI等待提示与终端侧边栏重复问题
- 修复模型调用工具后刷新页面,AI回复头部等待提示文字再次出现的bug
  - restoreTaskState 根据 hasAssistantContentEvent 精确设置 awaitingFirstContent
  - handleAiMessageStart 从头重建时保留 restoreTaskState 的设置
  - handleToolStart 兜底清除 awaitingFirstContent

- 修复刷新页面后终端侧边栏已有内容重复显示绿色箭头与输入内容的bug
  - 后端 get_snapshot/get_terminal_output 返回 last_event_time
  - 前端 TerminalPanel 基于 last_event_time 对 terminal_output 事件去重
2026-06-22 20:35:00 +08:00
f44c09a7ce fix(search): 移除 web_search 的 Tavily AI 摘要 2026-06-22 00:18:29 +08:00
b76708de13 fix(sub_agent): add missing datetime import in toolkit.py 2026-06-21 01:50:55 +08:00
f661dc6456 refactor(server,modules,utils,frontend): split oversized modules into packages and fix regressions
- Split server/chat.py, server/status.py, server/tasks.py into sub-packages
- Split utils/context_manager.py, utils/conversation_manager.py into mixin packages
- Split modules/file_manager.py, persistent_terminal.py, terminal_ops.py, mcp_client_manager.py into packages
- Split core/main_terminal_parts/context.py into base + mixins
- Split static/src/app/methods/{ui,message,upload,taskPolling,conversation} into sub-modules
- Fix flattened method exports, dynamic import paths, missing cross-module imports
- Add missing permission/network/execution mode constants and _PERMISSION_MODE_LABEL
- Update AGENTS.md, CLAUDE.md and project memories to reflect new structure
2026-06-20 15:54:48 +08:00
7b683ab390 fix(sub_agent): export TERMINAL_STATUSES in package __init__
fix(api_client): classify httpx.RemoteProtocolError as connection_error
2026-06-20 12:15:18 +08:00
9ed956518c refactor(sub_agent): 子智能体从 Node.js 子进程改为主进程内 Python 协程
- 重写子智能体执行核心,不再启动 easyagent Node.js 子进程
- 新增 modules/sub_agent/ 包集中管理子智能体逻辑
- 工具调用复用主进程 WebTerminal.handle_tool_call,自然经过沙箱/容器链路
- 子智能体模型独立读取 ~/.agents/<mode>/config/sub_agent_models.json
- 支持 8 个工具:read_file/write_file/edit_file(replacements+replace_all)/run_command/web_search/extract_webpage/search_workspace/read_mediafile
- 修复子智能体进度弹窗:标题颜色、write_file 显示、过滤非 progress 条目、统一滚动条样式
- 更新 AGENTS.md / CLAUDE.md 子智能体描述
- 新增 test/test_sub_agent_regression.py 回归测试
2026-06-20 00:26:45 +08:00
a212a0076c feat: support setting default workspace/project from frontend
- Add set_default_host_workspace() and /api/host/workspaces/set-default
- Add default_workspace_id persistence in UserManager and /api/projects/set-default
- Add 'Set as default' button in HostWorkspaceManageDialog
- Sync default badge in LeftPanel workspace list
- Update all workspace/project APIs to return real default_workspace_id
2026-06-19 21:48:24 +08:00
2766a8ab57 fix(sub-agent): 修复网络受限时无法创建子智能体,并修复详情页只显示最近一次工具调用的问题 2026-06-19 20:23:44 +08:00
9e4b0bdb97 fix: allow WebTerminal init without OCR API key (openai>=2.40 requires key) 2026-06-19 01:35:39 +08:00
c2d460a706 feat: 宿主机模式网络沙箱
- macOS sandbox-exec profile 支持 network_permission 参数 (restricted/full/none)
- backend: 透传网络权限至 run_command / terminal_session / sub_agent / background_command
- 后台 run_command 接入沙箱执行
- 自动审核模式兼容网络权限报错 markers
- 运行时切换网络权限通过 pending 机制 + user 消息通知
- 提示词注入网络状态 (仅沙箱模式)
- 前端权限菜单新增网络权限组 (受限/完全开放)
- direct 模式下网络权限组变灰禁用
- settings.json 默认 HOST_SANDBOX_NETWORK_PERMISSION=restricted
2026-06-19 00:22:30 +08:00
f31f8d6fd7 feat(host): host 模式同时加载 host/ 和 web/ 用户数据
- paths.py: 新增 IS_HOST_MODE / WEB_DATA_DIR / WEB_USER_SPACE_DIR
- user_manager.py: host 模式合并两处用户列表与工作区
- 已有 web 工作区直接复用,新数据写入 host/
- 更新 AGENTS.md / CLAUDE.md 文档
2026-06-16 22:39:32 +08:00
f43025047e refactor(config): 统一配置源,迁移数据到 ~/.agents/agents/
- config/paths.py: data_root 改为 ~/.agents/agents,config/ 提至根目录共享
- config/__init__.py: 统一加载 .env + settings.json,注入 legacy 环境变量
- 移除 4 处独立 .env 读取 (auth/model_profiles/api_client/chat_flow/balance)
- 新增 docs/config_refactor_plan.md 完整改造计划
- 数据已从 ~/.agents/host/ 迁移至 ~/.agents/agents/
2026-06-13 21:37:02 +08:00
dcb25dfe89 fix: 堆叠块深色hover不可见 & 底部块高度偏差 & 新增隐藏边线开关
- 深色模式 hover 修复:用 !important 覆盖 _overlays 中的 surface-panel 规则,使用 --hover-bg(6%白) 对齐对话列表 hover
- 底部块高度修正:shellHeight = targetShell + 2 补偿 border-box 下内容区裁剪
- 深色模式运行横条改为白色(var(--text-primary))
- 新增隐藏边线开关:后端 PersonalForm 持久化 stacked_hide_borders,前端条件渲染开关(仅堆叠模式可见)
2026-06-10 19:18:29 +08:00
3abbf30422 feat(admin): 管理员密码管理功能,支持搜索用户名/邮箱重置密码 2026-06-10 17:41:21 +08:00
4004bceb0c feat(config): 新用户默认deep模式+自动深压缩开启 2026-06-10 16:41:45 +08:00
960d087db0 fix(sub-agent): stdout output, real-time progress, config externalization
- config/sub_agent: 子智能体模型配置走 deploy_config_path 直指,无回退
- easyagent/config: ensureConfig 文件不存在时 throw 而非自动创建
- easyagent/batch: 所有输出走 stdout JSONL (emit),不写文件;conversation 走 emit 不写工作区
- sub_agent_manager: 新增 conversation_file;运行期间 select 非阻塞 drain stdout 即时写 progress
2026-06-10 03:09:31 +08:00
78544cb205 refactor: remove run_python tool, consolidate all execution to run_command 2026-06-07 16:23:01 +08:00
8617304fd4 refactor(prompt): freeze system prompts and update personalization modes
Freeze all leading system prompt blocks per conversation, including the main prompt, skills, workspace context, recent conversations, personalization, AGENTS.md, custom system prompts, and disabled-tool notices. Tool schemas remain dynamically generated outside the frozen prompt path.

Remove obsolete prompt injection files and code paths for thinking mode, deep thinking mode, sub-agent guidelines, and todo guidelines so prompt assembly stays simpler and cache-friendly.

Add personalization-controlled conversation_continuity with high/medium/low levels, rename the UI copy to 对话连续性, and wire it through backend sanitization, prompt generation, manage_personalization, frontend settings, and tool result rendering.

Add auto communication style support via prompts/auto_style.txt, matching the existing separate-file pattern used by human_like_style.txt.
2026-06-07 02:52:00 +08:00
9be775d346 feat(personalization): 简略消息显示改为持久化到配置文件
原来「简略消息显示」只存浏览器 localStorage(experiments),换设备/清缓存即丢。
现迁移到后端 personalization.json,与主题/默认隐藏工作区一致即时保存:

- 后端新增 compact_message_display 字段(默认 full)及清洗(仅 full/brief,
  容错大小写空格,非法回退 full)
- store: PersonalForm 加字段;setCompactMessageDisplay 写 form + 即时 POST;
  加载时同步进旧 experiments 镜像,并一次性迁移本地遗留 brief 回写配置文件
- 抽屉下拉与 ChatArea 改读 form.compact_message_display,experiments 作兜底
- 保留 localStorage 镜像保证向后兼容,老用户首次加载自动迁移偏好

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-05 23:31:06 +08:00
b8a51c1c63 refactor(config): 移除硬编码模型残留,部署级配置外置到 ~/.agents
## 模型逻辑清理
早期把模型 API 端点/密钥/模型 ID 硬编码的残留(AGENT_API_* / THINKING_* /
TITLE_* 三件套)已彻底移除。模型配置统一由 config/custom_models.json
(经 model_profiles 解析为档案)描述,运行时通过 apply_profile 注入;
没有任何可用模型时按既定行为报错。

- config/api.py: 删 9 个三件套符号,仅保留 DEFAULT_RESPONSE_MAX_TOKENS
- utils/api_client.py: client 改为空配置起步,移除三件套 import
- server/chat_flow_helpers.py: 删死 import(TITLE_* 符号)

## 部署级配置外置(按"决策权归谁"分类)
config/*.json 不再一概锚定源码树:
- 程序能力(docker_risk_markers / skill_hints)留源码树,随版本演进
- 部署者自定义(custom_models / host_workspaces / auto_approval /
  goal_review / forbidden_commands / host_sandbox_policy)外置到
  ~/.agents/<mode>/config/

新增统一解析机制(config/paths.py):
- DEPLOY_CONFIG_DIR(默认 ~/.agents/<mode>/config,可用环境变量覆盖)
- resolve_deploy_config():只读,回退链 部署目录→源码树.json→源码树.example
  (开发环境不必先跑 setup 也能用源码树种子)
- deploy_config_path():写回用,稳定指向部署目录

6 个加载点改用上述解析;顺带修复 approval_agent / goal_review_agent 的
DEBUG_TRANSCRIPT_DIR 仍指旧源码树 logs/ 的漏迁问题。

## 安全与运维
- 含密钥/机器特定的 5 个文件停止 git 追踪(git rm --cached,本地保留)
  并加入 .gitignore,仓库仅留 .example 种子;forbidden_commands 保留
  追踪作默认黑名单
- scripts/setup.py: 模型配置写到部署目录(用与 paths 一致的独立推导,
  不 import config 以免过早锁定路径)
- scripts/migrate_runtime_data.py: 新增 config/*.json → 部署目录迁移
  (备份 + 不覆盖已存在)

## 关联:P1 配置收敛 + P2 首启向导
- config/server.py: Web 端口/监听地址/debug/NODE_BIN/PYTHON_BIN 单一事实源
- 消灭 8091 多处重复定义(state.py 死代码、app_legacy、main.py 各读各的)
- 修复 sub_agent_manager 命令数组硬编码 "node" → NODE_BIN(便携包内置
  Node 的前提)
- scripts/setup.py: 终端首启向导(模式/端口/管理员/密钥/模型)

## 测试
test_config_paths_resolution 更新以反映新行为(host_workspaces 锚定部署
目录、新增 DEPLOY_CONFIG_DIR 用例);全部离线用例通过。

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-01 16:59:34 +08:00
6f78277a03 feat(ui): add project git status panel 2026-05-31 17:27:24 +08:00
6d5630c820 feat(file-edit): support batched replacements 2026-05-31 13:15:37 +08:00
3fd2214da0 feat(compression): 深压缩改为 in-place 标记前缀,不再切换新对话
深压缩从"建新对话+切过去"重构为标记当前对话历史前缀(deep_compacted):
原文保留并照常显示,仅在 build_messages 构建请求时排除。conversation_id
不变,避免任务状态/目标模式/前端对话切换的大量适配带来的 bug。

- deep_compression.py: _mark_history_compacted 打标 + 重置 current_context_tokens
  防自动续接死循环;总结请求传入正常 tools 以 100% 命中前缀缓存
- context.py build_messages: 跳过 deep_compacted 消息
- conversation.py /compress: 去掉切对话,按 compress_behavior 决定续接/等待
- 新增个性化设置 deep_compress_form(file/inject) 与
  deep_compress_behavior(continue/wait,仅手动压缩生效)
- 前端去掉强制切换对话,改为重载当前对话刷新展示
- AGENTS.md/CLAUDE.md: 补充默认中文交流约定等说明

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-05-30 20:56:15 +08:00
8a1e17b9a0 feat(goal): add autonomous goal mode workflow
Introduce workspace-level goal state persistence, goal prompt injection, and after-turn review handling so an active task can continue until the configured completion conditions are met.

Add a dedicated goal review agent with readonly and active evidence modes, configurable model settings, review prompt, token/turn boundaries, idle-no-tool protection, and progress/completed/stopped events.

Wire goal_mode through task creation, task restoration, compression handoff, runtime user messages, API message sanitization, and tool-call ordering so goal continuations survive long-running tasks and deep compression.

Add Vue UI for arming goal mode from the quick menu, showing running/completed banners, displaying progress metrics, restoring running goal state, and exposing personalization settings for review mode and stop limits.

Include goal mode research notes and default goal review configuration.
2026-05-30 12:51:42 +08:00
08d9812f45 feat(docker): support isolated projects in web mode
Implement Docker Web project management on top of workspace_id while keeping host mode workspace behavior intact.

Backend changes:

- Move Docker user files to users/<user>/projects/<project_id>/{project,data,logs} and shared user state to users/<user>/personal.

- Add automatic legacy migration from old project/data/logs/agentskills layouts and from the temporary project/shared layout.

- Generate project IDs server-side from the project name, reserve internal IDs, and hide project paths from the UI.

- Add project rename/delete APIs; deleting a Docker project removes its project directory and releases terminal/container state.

- Add host workspace rename/delete APIs; deleting a host workspace only removes catalog entries and keeps files on disk.

- Scope Docker containers, terminals, uploads, conversations, storage, and running task metadata by workspace_id.

- Store private skills under users/<user>/personal/agentskills and sync them into each project workspace.

- Inject the current project/workspace name through prompts/workspace_system.txt.

Frontend changes:

- Replace Docker file-tree area with project management UX and reuse host multi-workspace running-task behavior.

- Add a management dialog for creating, renaming, and deleting projects/workspaces.

- Show project labels instead of internal IDs in toasts and project lists.

- Clear stale conversation history during project/workspace switching and show loading until the new list is ready.

- Keep running-task loader/check behavior consistent across host workspaces and Docker projects.

Validation:

- python3 -m py_compile modules/user_manager.py modules/host_workspace_manager.py modules/skills_manager.py core/main_terminal_parts/tools_execution.py server/auth.py server/context.py server/status.py server/conversation.py server/tasks.py utils/context_manager.py

- python3 -m unittest test.test_server_refactor_smoke

- npm run build --silent 2>&1 | tail -n 20
2026-05-29 19:49:48 +08:00
eb00a3522a refactor(models): use dynamic api model registry 2026-05-29 00:20:54 +08:00
f61a8df6fa feat: add blocking user question tool 2026-05-28 16:30:55 +08:00
1812ee7c27 fix(skills): migrate read_skill path to .agents/skills 2026-05-27 19:17:53 +08:00
398ee6e760 feat(memory): refactor proactive memory and conversation recall
Refactor conversation persistence into workspace-scoped storage with host workspace legacy migration, current workspace filtering, and active host workspace state protection.

Add proactive recent-conversation prompt injection with personal-space controls, configurable item count, empty-conversation filtering, and frozen per-conversation prompt metadata to keep request caches stable.

Introduce conversation_search and conversation_review tools for current-workspace history: multi-keyword search, list mode, current conversation exclusion, message/tool counts, formatted model output, read/save review modes, long-review fallback, and .agents/review storage.

Move internal workspace artifacts under .agents, including uploads, skills, compact results, and reviews; update prompt file-tree visibility, compression output, upload paths, and hidden versioning exclusions.

Add create_skill workflow and a Skills tool category, move read_skill into Skills, validate/归档 skill folders, support host global skills and Docker/API private skills, and sync private skills into .agents/skills and skill prompts.

Update frontend tool icons, statuses, and expanded renderers for conversation recall and skills workflows; add workspace conversation and skills-manager tests.
2026-05-27 03:09:42 +08:00
16473c824d chore: 合并多项独立 WIP 改动
- cli: 增加 layoutRefreshTick 与 IME 输入组件,处理 slash_menu 光标偏移
- config/auto_approval: 切到 opencode 中转端点
- modules/approval_agent: 调试 transcript 增加 trace / final_result 字段
- static/useLegacySocket: 重试 toast 显示具体错误信息

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 02:01:36 +08:00
67fdb715f5 refactor(messages): 统一运行期消息插入为 [系统通知|source] 前缀的 user 消息
- 新增 server/chat_flow_task_support.inject_runtime_user_message 作为唯一入口,
  统一处理 add_conversation + messages.insert + sender(user_message) 三件事
- 子智能体 / 后台 run_command / runtime_guidance / mode notice 全部走该入口,
  role 一律 user,inline 仅作 metadata 标记,不再复用 system role 表达语义
- inline poll 移出 tool 执行循环,改为整轮 tool 完成后批量注入,
  避免在 assistant.tool_calls 与 tool 序列之间插入 user 导致 API 报错
- 删除 _insert_completion_notice_message / _record_sub_agent_message
  以及 build_messages 中 system→user 的回转分支
- 移除已废弃的 wait_sub_agent 工具残留(后端注册、前端图标/动画、
  prompts、SKILL 文档;保留 sleep 的 wait_sub_agent_ids 参数)
- 前端取消对 is_auto_generated/auto_message_type 的过滤,
  让运行期注入的 user 消息在历史和实时推送中正常渲染

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-23 02:01:36 +08:00