Enhance platform sync management and add new configurations - #19
Merged
Conversation
- Introduced Cline platform configuration in `env/platforms/cline.json`. - Implemented syncing of global state and secrets for Cline in `sync/platforms/cline.py`. - Updated README and documentation to reflect Cline integration. - Enhanced sync process to skip platforms not installed, including Cline, Codex, Claude, CodeBuddy, Gemini, and Continue. - Added tests to ensure proper handling of missing platform directories and configurations. - Refactored sync logic to improve clarity and maintainability across platforms.
- Updated README to reflect the addition of Qwen Code with 9+ AI coding tools. - Added Qwen Code configuration in `env/secrets.json.example` for API key management. - Enhanced sync process to include Qwen Code, updating paths and syncing skills. - Modified TypeScript configuration to support Qwen Code in the CLI. - Improved CLI commands for better usability and error handling.
- Revised Cline configuration paths to include specific files for better clarity. - Updated Gemini CLI documentation to specify relevant configuration files. - Added Qwen Code configuration details to the sync output for improved user guidance.
…nt during sync - Added a sidecar to track keys managed by Cline, allowing for automatic pruning of stale keys from globalState.json and secrets.json when the platform is disabled. - Updated sync logic to remove unmanaged keys, enhancing the cleanup process for disabled platforms. - Revised documentation to reflect changes in key management and sync behavior.
- Changed actModeOpenAiModelId in cline.json from deepseek-v4-pro to deepseek-v4-flash for improved model performance.
- Updated the comment in cline.json to provide a clearer explanation of the global state and secrets sync process, emphasizing the merging of keys and the handling of placeholder values.
… skill bundle integrity checks - Updated `validate.sh` to skip links pointing outside the skill root directory, ensuring only valid internal links are checked. - Enhanced `skill_bundles.sh` to copy necessary directories and files for skill distribution, including a whitelist for allowed directories. - Improved `validate-skill-integrity.sh` to assert the existence of `SKILL.md` in single skill mode, preventing silent passes on integrity checks. - Added validation for optional MCP definitions in `optional_mcps.sh` to ensure only valid configurations are enabled and synced.
…kill proposal script - Updated comment to specify that the active version of the ledger remains unchanged while only draft proposals are created and registry deduplication occurs.
…_skills function - Added a check to ensure that the wait command for background process IDs only executes if there are active processes, preventing unnecessary errors when no processes are running.
…nagement - Introduced `sync-memory.sh` to facilitate the accumulation and retrieval of user memory across sessions. - The script allows users to inject managed blocks into various target files, ensuring persistent memory storage in `~/.ai-coding-kit/MEMORY.md`. - Added functionality for remembering and recalling memories with optional tagging, enhancing user interaction and memory management.
…zation - Added `sync-user-profile.sh` to manage user profiles across sessions, injecting a `user-profile` block into preambles. - Introduced `sync-memory.sh` for event-level memory management, allowing users to remember and recall memories across sessions. - Updated `bootstrap.sh` and `run-sync.sh` to include these new scripts, enhancing the synchronization capabilities of the system. - Documented new features in the README, including options to skip user profile and memory synchronization during setup.
…ate README - Introduced `review.json` for configuring auto-code-review parameters, including options for enabling the feature, specifying reviewers, and setting review limits. - Updated `README.md` to document the new `review.json` file and its usage, along with details on optional MCP servers and their management. - Added `review.json.example` to provide a template for users to create their own configuration.
…latform support - Modified the command execution in `lanhu.json` to use `exec` for better script handling. - Added `codebuddy` to the list of supported platforms, enhancing compatibility.
…mble.sh - Modified the skills directory path to replace the home directory with a tilde (~) for better readability and compatibility in the script.
- Updated paths in various scripts to reflect the new directory structure, ensuring compatibility and clarity. - Modified `.gitignore` to include `__init__.py` files for better package management. - Enhanced documentation in `README.md` and `model_routing.md` to reflect changes in script locations and usage. - Removed deprecated `sync_config.py` and related files, streamlining the sync process.
…pt paths - Added support for backup configuration through `env/backup.json`, allowing users to specify a custom backup directory. - Updated script paths in various locations to reflect the new directory structure, ensuring consistency across the project. - Removed deprecated scripts from the `sync` directory, streamlining the codebase. - Enhanced documentation in `README.md` and related files to reflect changes in backup functionality and script usage.
- Deleted the `model_routing.md` file, which contained design specifications for a unified provider layer for multi-platform model routing. This change reflects a shift in project direction and the need to streamline documentation.
…orms - Added `api.enabled` configuration to control synchronization of models and secrets for Continue and Cline platforms. - Updated documentation to clarify the behavior of the `api.enabled` toggle and its impact on managed fields. - Enhanced sync logic to ensure proper cleanup of managed keys when API sync is disabled. - Introduced tests to validate the new synchronization features and their behavior under different API sync states.
- Introduced `api.enabled` configuration for Codex platform to control synchronization of API fields and environment variables. - Updated documentation to clarify the behavior of the `api.enabled` toggle and its implications for managed fields. - Enhanced sync logic to ensure proper deletion of API-related keys when API sync is disabled, adhering to the cleanup policy. - Added tests to validate the functionality of the new API sync features and their behavior under various states of the `api.enabled` toggle.
- Updated `api.enabled` setting to false, disabling API synchronization by default. - Adjusted comments in the configuration file to reflect the change in API sync behavior.
- Updated the OUT-OF-SCOPE documentation to specify that certain scenarios are not processed by the skill and do not trigger automatic review. - Enhanced the auto-code-review documentation with detailed triggering logic, outlining explicit user actions required to initiate a review.
- Updated the test configuration to explicitly set `api.enabled` to True, preventing tests from inheriting the value from the codex.json file. - Added comments to clarify the purpose of this change and guide future test configurations regarding the `api.enabled` toggle.
- sync/platforms/codebuddy.py: preamble.mode=full 时不再双写独立 historical-recall 块，改由 sync-agent-preamble.sh 渲染内嵌 recall 的完整 preamble；recall 仍写独立块，none 跳过 - tests/test_registry.py: codebuddy 由 recall 列表移入 full 列表 - tests/test_codebuddy_sync.py: 默认配置对齐真实 mode=full，新增 full/none 三态断言，保留 recall 路径覆盖 - docs: 同步 platform-sync-contract.md 与 sync/README.md 的 CodeBuddy 描述 - skills-engineering/ios-engineer/SKILL.md: 扩写 description，覆盖多 Apple 平台 (iPadOS/macOS/watchOS/tvOS)、Objective-C(++)/Combine/async-await/ WidgetKit/App Extensions 与中文诊断术语；附带 evolution proposal + approval (20260723-152751-expand-description-platforms)
将 managed-block 标记从 `ios-engineer` 重命名为更具通用性的 `agent-preamble`，并兼容迁移旧标记；同时重构模板以消除历史召回段落的重复。
- D1: CAM 激活时抑制 preamble 轻量校准段，Tier0/Tier2 互斥扩展到 preamble 层 - D2: GR-002 前置确认被 PG-000 盘问吸收、GR-006 中断与 GR-002 同 anchor 合并 - D3: 新增多 SKILL 叠加总纲与分级读取/预算上限，补行为测试防回归 - D4: GR-004 与 CAM 详规对齐——不重复语义但保留 CAM 机械格式（ios-engineer 提案 20260723-173058-cam-fields-preserve-format） - D5: 跨块置信度归一到本轮唯一保留的字段 - 同步 engineering-discipline/plan-grill/ios-engineer/cognitive-expansion 的 en-US 镜像
- 新增 tests/test_en_us_mirror_sync.py：锁定 engineering-discipline/plan-grill/ios-engineer/cognitive-expansion 的 zh 源与 en-US 镜像双向协同条款锚点，防 en-US 再次静默滞后\n- CHANGELOG 3.0.0 段补 Changed(D1-D5 多 skill 协调) 与 Fixed(en-US 分发闭环) 记录
- 新增 skills-engineering/scripts/validate-global-skills.sh：只读验收入口，串起结构/行为/preamble dry-run/同步验证/integrity --check-only/全局协调回归测试\n- tests/test_codebuddy_sync.py 补 GlobalSkillValidationScriptTests 校验该入口只读且覆盖闭环步骤\n- README 说明本地一键跑完整验收闭环
- 移除 20260608-add-engineering-quality-gates 的 proposal/approval/validation 三份记录（超出设计保留的最近 10 份窗口）
- 变更记录补 3.0.3：多技能叠加协调、en-US 镜像闭环、回归护栏、演进记录 10 份保留策略\n- 校验与观测新增「全局协调与 i18n 回归测试」小节，说明 test_en_us_mirror_sync.py 与 validate-global-skills.sh\n- 开发建议补 en-US 同步纪律：改 zh 源须同步镜像否则回归测试红
- 顶层沉淀 D1-D5 多技能叠加协调契约（GR-002↔PG-000 吸收、GR-006 同 anchor 合并、GR-004↔CAM 格式保留、跨块置信度归一、分级读取预算）\n- i18n 治理：zh 源 + en-US 镜像纪律与覆盖校验
- 该文档对应的同步架构优化已收尾，留痕于 CHANGELOG 与 skill references，故移除临时优化文档
- 协调契约引导句补全 CAM 括号称谓并交叉引用 .agents/composition.md（块发射顺序与冲突裁决）\n- i18n 治理统一「分发镜像」术语，与全文 en-US 镜像表述一致
…on local-only
- 目录重命名 env/optional-mcps/ -> env/optional_mcps/，与 optional_mcps.sh 及 env 下其它子目录（无连字符）命名一致
- 修正 optional_mcps README 用法示例 playwright -> puppeteer（puppeteer 为默认 env/mcp/playwright.json 的可选替代）
- enabled.json 不再提交：加入 .gitignore 并从索引移除（磁盘保留，脚本缺失时自动重建为 {}）
- 同步更新脚本、schema 校验、env/README、sync/README、CHANGELOG 的引用
Refactor sync logic and enhance API management across platforms
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.