mirror of
https://github.com/agentscope-ai/ReMe.git
synced 2026-09-06 08:16:00 +00:00
3 commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
jinliyl
|
83831ec90c
|
feat(core): enhance reme4 (#281)
### 1. Agent Wrapper（统一 Agent 后端抽象） - **`base_agent_wrapper.py`**：`reply()` 返回值从 `tuple[str, Any]` 改为 `dict`（含 `session_id` / `last_message` / `result` / 可选 `structured_output`）；`reply_stream()` 改为产出统一的 `StreamChunk`。废弃 `add_tools()`，改为 `add_job_tools(names: list[str])`（按名解析 BaseJob）与 `add_skills()`；新增 `_resolve_job_tools()`、`_merged_kwargs()`、`_chunk()` 辅助方法及 `project_path` / `project_skills_root` 属性。 - **`as_agent_wrapper.py`（AgentScope 后端）**： - 会话持久化重写：`session_path` 落地到 `<vault>/<session_dir>/agentscope/`，`_load_state` 支持 `resume` / `session_id` / `fork_session`，并做 UUID 校验（`_validate_session_id`）；`_cleanup_expired_sessions` 按天数清理过期会话。 - 新增内置工具集（`BypassAnalysisBash` + Edit/Glob/Grep/Read/Write），`BypassAnalysisBash` 绕过 AgentScope 自带 Bash 静态分析以让 permission_mode 生效；`_resolve_skills()` 把配置的 skill 暴露给后端，`_load_tool_env()` 注入项目 `.env`。 - `_event_to_chunk()` 把 20+ 种 AgentScope 事件（Reply/Text/Thinking/Data/ToolCall/ToolResult/ModelCall/ExceedMaxIters）归一化为 `StreamChunk`。 - **`cc_agent_wrapper.py`（Claude Code SDK 后端，+551 行）**： - 新增 `_CcFileSessionStore`：基于 vault 的文件型会话存储，实现 append（按 uuid 去重）/ load / list / delete / list_subkeys，并对路径做 `_safe_parts` + `resolve()` 防越界校验。 - `_build_options()`：统一构建 `ClaudeAgentOptions`，处理 skills、disallowed_tools（默认禁 `WebSearch`）、`.env` 注入、Claude Code 的 API 凭据解析（`_claude_code_api_env`，多级 base_url/api_key 回退）、`CLAUDE_CONFIG_DIR` 设置、skill 目录软链接（`_ensure_claude_skill_dir`）。 - `_raw_event_to_chunk()` / `_message_content_to_chunks()`：把 Anthropic 流式事件（message_start/delta/stop、content_block_*）与 SDK 消息块（AssistantMessage/UserMessage/ResultMessage/RateLimitEvent）转换为统一 `StreamChunk`；跟踪 block_id/block_type/tool_call_name 做关联；处理尾部 `"success"` 误报异常的吞掉逻辑。 ### 2. 统一流式协议（StreamChunk / ChunkEnum） - **`stream_chunk.py`**：`StreamChunk` 扩展为承载 AS + CC 双后端完整信息的统一结构，新增 `session_id` / `block_id` / `tool_call_id` / `tool_call_name` / `media_type` / `input_tokens` / `output_tokens` 等字段，纯文本流仍保持轻量。 - **`chunk_enum.py`**：补全生命周期标记 `REPLY_START` / `REPLY_END`，并文档化两套后端事件 → ChunkEnum 的映射。 ### 3. Index 模块重构（变化批次化 + dispatch） - 新增 `_change_batch.py`：`coalesce_changes()` 把同路径多次事件折叠为最终状态（结合 path 存在性判定），`bucket_changes()` 按 watchfiles.Change 分桶。 - 新增 `init_changes.py`（`InitChangesStep`）：一次性扫描，对比 file_store / file_catalog 已索引节点计算 added/modified/deleted，写入 `context["changes"]` 后 dispatch。 - 新增 `update_changes.py`：抽象基类 `ChangeApplyStep` 统一 added/modified/deleted 处理与错误收集；`UpdateCatalogStep`（写 file_catalog）、`UpdateIndexStep`（写 file_store，含按后缀解析 chunker）。 - **`watch_changes.py`**：改用 `dispatch_step_specs`（基类提供的 `dispatch_steps()`），每批先 `coalesce_changes` 再 dispatch；默认参数调整（debounce 5000ms / step 1000ms / poll 5000ms）并暴露常量。 - 删除旧步骤：`clear_and_scan` / `foreach_dispatch` / `scan_changes` / `update_catalog`(旧) / `update_index`(旧)；`clear_store.py` 取代 clear_and_scan。 ### 4. Evolve / Dream 模块（拆分为多步 pipeline） - 删除旧的单体 `auto_dream.py` / `dream.py` / `dream.yaml`，新增 `dream/` 子包，按 5 个步骤组织： - **`extract.py`**：扫描当日 day-index + daily 笔记，对比 file_catalog 找出 changed/deleted，调用 LLM 全局抽取 `units`（procedure/personal/wiki 三桶）与 `topics`，路径与桶做清洗/路由。 - **`integrate.py`**：逐个 unit 调用 LLM 写入 digest，结构化输出 `IntegrateOutcome`（CREATE/CORROBORATE/REFINE/CORRECT），失败 unit/路径收集回写。 - **`topics.py`**：写 `daily/<date>/interests.yaml`，结合当天已有 + 近 N 天做去重（`normalize_topic`），可走 LLM 或纯规则去重两条路径。 - **`proactive.py`**：读取当日 `interests.yaml`，作为主动推荐话题的入口。 - **`finish.py`**：把变更路径落盘到 dream file_catalog（checkpoint），渲染最终汇总摘要。 - 新增 `schema.py`（`DreamState` 等跨步骤共享状态与结构化输出模型）与 `utils.py`（状态存取、扫描打包、YAML 读写、结构化回复解析等公共函数）。 - `evolve/__init__.py` 导出全部新 step。 ### 5. auto_memory / auto_resource（适配新 Agent API） - **`auto_memory.py`**：会话路径迁移到 `<session_dir>/dialog/<session_id>.jsonl`；改用 `job_tools`；新增 `source_conversation` frontmatter 反向链接（`_session_link`）；执行后刷新 day 索引（`refresh_day_index`），并对 session_id 做合法性校验。 - **`auto_resource.py`**：资源改用「同名 daily note」方案（`_compute_note_stem` 取文件 stem）；批量处理 `changes: list[dict]`（`_handle_change` 逐项处理，返回逐项结果摘要）；agent 会话 id 用稳定的 `uuid5`；同样刷新 day 索引。 ### 6. BaseStep 基类增强 - 新增 `dispatch_steps` / `dispatch_step_specs` 机制：`_resolve_dispatch_step()` 支持字符串或 dict 形式的 step spec，`dispatch_steps()` 复用当前 context 调用下游 step。 - 新增 `config_value()`：按 key 取 app config，缺失时回退 `ApplicationConfig` 默认值。 - 小幅清理：`language` 初始化、`copy()`、`Ref.__init__` 签名精简。 ### 7. Components 改动 - **`file_store/local_file_store.py`**：持久化改用 zstd 压缩（`.jsonl.zst`，通过新 `utils/jsonl_zst.py`）；upsert 时先删除旧 chunk 的 keyword 文档；embedding 复用改为 `(text, embedding)` 键控，要求文本一致才复用；新增 `_matches_search_filter()` 对 vector/keyword 搜索做 path/path_prefix/metadata 的统一后过滤。 - **`keyword_index/bm25_index.py`**：索引文件名加入组件名 + tokenizer 指纹（sha256 前 12 位），快照/恢复时校验指纹防配置漂移；空索引 dump 时删除文件，加载失败抛错而非静默。 - **`file_chunker/markdown_file_chunker.py`**：弃用 `python-frontmatter`，改用内置 YAML 解析（非法 YAML 不阻断正文索引），并修正因 frontmatter 占用行号导致的 AST 行号偏移（`line_offset`）。 - **`cron_job.py`**：大幅简化（-187 行），由原来「dispatch 外部 job/step + 多种调度模式」改为「在自身 steps 上跑 cron 表达式」；`Application` 启动顺序随之调整为 base > stream > background > cron。 - 其余小调整：service（base/http/mcp）、file_graph、file_catalog、as_llm、as_embedding、tokenizer、prompt_handler、base_component 的签名/接口微调。 ### 8. Application 生命周期 - `_start()` 启动顺序明确为 components → base → stream → background → cron，启动失败会触发 `_close()` 回滚并 re-raise（不再吞异常）。 - 启动时创建 `session_dir` 目录；新增 `update_component()`（按类型/名就地更新已存在组件，不存在则报错）。 ### 9. File IO / 路径安全 - **`_path.py`**：`resolve_path` 增加 vault 越界防护（`is_relative_to` 校验），禁止 `.` / `..` 路径分量，支持 `allow_empty`。 - **`read.py`**：大文件（超过 `MAX_FILE_READ_BYTES`）走按行读取 `read_file_lines_safe`，避免一次性载入内存。 - **`_file_io.py` / `_daily_index.py` / `_path.py`** 等支持函数补齐（如 `refresh_day_index`、`read_file_lines_safe`）。 - **`env_utils.py`**：新增 `parse_env_file()`，`load_env()` 返回加载到的键值、支持 `override`、对无路径调用做幂等缓存。 ### 10. Config - `ApplicationConfig` 新增 `session_dir`（默认 `reme_session`）。 - `config_parser.py`：环境变量展开后做类型转换（`_convert_value`）、dot-notation 与 key=value 参数校验更严格、配置文件路径支持相对 `_CONFIG_DIR` 查找、根非 dict 报错。 - `default.yaml`：作业编排改用 `init_changes_step` + `dispatch_steps`（index/resource/digest 三个 watch loop 与 reindex）；新增 `auto_dream`（4 步）、`proactive` 作业，移除旧 `dream`；file_catalog 增配 `resource` / `digest` / `dream` 实例；LLM 默认值与 Claude Code 凭据配置调整（tool_result_limit 50000、thinking_enable=false 等）。 ### 11. 其它 - 新增 `steps/common/add.py`（`AddStep` 算术 demo）、`channel/__init__.py` 与 common `__init__` 导出整理。 - 新增 4 篇文档：`docs4/auto_dream_logic_and_step_refactor.md`、`docs4/watch_loop_step_refactor_plan.md`、`docs4/todo.md`，以及 `reme_design.md` 更新。 ** |
||
|
jinliyl
|
d8086039dc
|
refactor(Agentscope2.0): llm & embedding & agent (#271)
* refactor(embedding): replace embedding model with embedding store architecture - Remove as_token_counter component and its estimated token counter implementation - Replace BaseEmbeddingModel with BaseEmbedding that wraps AgentScope embedding models - Add support for multiple embedding providers (OpenAI, DashScope, Gemini, Ollama) - Introduce BaseEmbeddingStore and LocalEmbeddingStore for caching and persistence - Update component registry to use new embedding and embedding_store types - Modify file stores to use embedding_store instead of embedding_model - Update health check to monitor embedding_store instead of embedding_model - Change default config to use embedding_store with local backend - Add estimate_token_count utility function to utils module * refactor(llm): replace as_llm components with unified llm implementation - Remove deprecated as_llm and as_llm_formatter modules - Add new llm module with BaseLLM and provider-specific implementations - Update component registry to use LLM instead of AS_LLM - Replace all as_llm/as_llm_formatter references with llm in steps - Update configuration schema to use llm instead of as_llm - Rename integration test file from test_as_llm to test_llm - Add proper docstrings to embedding store dimension property - Add pylint disable comment for embedding model call - Remove unused FormatterBase import in base_step - Update token_utils with function docstring * refactor(evolve): replace ReActAgent with Agent and update message handling - Removed FlexReActAgent class and direct ReActAgent imports - Updated Agent instantiation to use new constructor parameters - Changed message content to use TextBlock format instead of plain strings - Modified timestamp access from msg.timestamp to msg.created_at - Updated metadata access pattern for structured outputs - Replaced Msg.from_dict with Msg.model_validate in auto_memory.py - Updated test mocks to patch Agent instead of ReActAgent - Changed message serialization from to_dict to model_dump in tests - Moved component references to base class definition - Updated demo tools to return strings instead of ToolResponse objects * feat(step): migrate to FunctionTool and add streaming support - Replace deprecated ToolResponse with FunctionTool in base_step.py - Remove unused TextBlock import from base_step.py - Update job registration to use new FunctionTool API - Add thinking_budget parameter to llm_demo configuration - Introduce StreamLLMDemoStep with streaming output capability - Add structured output support to LLMDemoStep via generate_structured_output - Implement streaming event handling for text/thinking/tool calls - Add integration tests for embedding functionality - Add integration tests for structured output and streaming features - Update tool usage in demo steps to use new function naming convention * fix(ci): correct package installation path in unittest workflow - Updated pip install command to use proper package path "./reme4[dev,core]" - Fixed dependency installation step in CI workflow configuration * chore(workflow): update python versions in unittest workflow - Remove Python 3.10 from test matrix - Add Python 3.11 to test matrix - Add Python 3.12 to test matrix - Keep Python 3.13 in test matrix - Update matrix configuration for better version coverage * fix(health): handle missing dimensions attribute in embedding status - Wrap dimensions access in try-except to prevent AttributeError - Return None when dimensions attribute is not available - Maintain backward compatibility for components without dimensions test(component): add comprehensive tests for BaseComponent and related classes - Add tests for Dependency class including repr and attribute access - Add tests for bind method with various scenarios and edge cases - Add tests for lifecycle management and async context handling - Add tests for standalone and context-bound dependency resolution - Add tests for ComponentMixin path utilities test(common): update LocalFileStore initialization parameter - Change embedding_model parameter to embedding_store in test setup - Update all affected test files consistently test(registry): add complete test suite for ComponentRegistry - Add tests for register method with explicit names and defaults - Add tests for decorator registration pattern - Add tests for get_all method returning copies - Add tests for unregister and clear operations - Add tests for error handling of invalid registrations test(job): add comprehensive tests for BaseJob and BackgroundJob - Add tests for step resolution and exception handling - Add tests for backoff delay calculation with jitter - Add tests for supervisor loop restart behavior - Add tests for task shutdown and cancellation test(prompt): add complete test suite for PromptHandler - Add tests for prompt loading from dictionaries and files - Add tests for internationalization and language fallback - Add tests for flag filtering and variable substitution - Add tests for format validation and error handling test(runtime): add basic tests for RuntimeContext dictionary access - Add tests for item getting, setting and containment checks - Add tests for missing key error handling * feat(evolve): add permission context and agent state management - Import PermissionContext, PermissionMode and AgentState modules - Add state configuration with bypass permission mode to AutoDream agents - Add state configuration with bypass permission mode to AutoMemory agents - Implement static _to_msg method for message validation and formatting - Refactor message processing to use the new _to_msg method - Ensure proper content structure for text blocks in message conversion * style(tests): update test files with linting rules and code improvements - Add missing pylint disable directives for docstring and attribute warnings - Replace lambda expressions with proper function definitions in test cases - Import Path directly instead of using lambda with __import__ - Simplify assertion checks by using truthiness instead of equality to empty dict - Remove unused imports and reorder imports consistently - Format dictionary literals with proper indentation and line breaks |
||
|
jinliyl
|
3cb2579ff7
|
refactor(steps): update auto-memory (#263)
* refactor(steps): update naming conventions in components and configuration Updated naming conventions across multiple files, changing colon-separated names to underscore-separated format, and added new step definitions along with documentation updates. Key changes: - Replaced `Synchronizer` with `AutoMemory` as the counterpart component for cold-write operations - Updated naming conventions in all related configuration files (e.g., `frontmatter:read` → `frontmatter_read`) - Added new step definitions such as `submit_slug_updates` and `auto_memory` - Updated relevant documentation - Modified log output format for improved readability * refactor(evolve): Refactor the auto-memory module and update related configurations - Remove the old slug update commit step file - Add new auto-memory planner and writer steps - Update __init__.py to export the new step classes - Modify the auto_memory configuration structure in default.yaml - Update the slug field description for clearer explanation of its purpose * up * up * refactor(tests): Move unit test directory from `tests4/unittest` to `tests4/unit` Additionally, the assertion logic in test files has been updated: direct comparisons of `payload["notes"]` have been replaced with checks verifying the presence of paths and metadata within the response content. Furthermore, some test expectations have been simplified—for example, using `count` instead of asserting against specific note lists. Specific changes include: - Updating workflow configurations to align with the new test directory structure - Modifying assertions across multiple test methods to make them more flexible and maintainable - Cleaning up and optimizing parts of the test code structure This is a comprehensive test refactoring effort aimed at improving test readability and robustness. * Refactor(steps): Update memory writing logic and optimize JSON schema structure Improved the write strategy description in `auto_memory_writer.yaml` to emphasize using `edit` over `write`. Adjusted the `json_schema` structure in `base_step.py` to support the new function definition format. Also corrected grammatical issues in the related documentation. * Fix: Improve frontend data parsing error handling and update test files Added capture and handling logic for YAML parsing exceptions, providing more detailed error messages when frontend data format issues occur. Also corrected the description text in a test file. |
Renamed from tests4/unittest/test_common_steps.py (Browse further)