feat(automations): 条件驱动的自动化工作流 - #94
Merged
Merged
Conversation
Keep one-way startup migration from scheduled_tasks.json, and drop the old /schedules API plus scheduler.* tools so only automations remain as the entry. Co-authored-by: Cursor <cursoragent@cursor.com>
Move crontab add/update and ScheduledTask storage out of the runtime so time jobs only carry task_id and workflows run through the new service. Co-authored-by: Cursor <cursoragent@cursor.com>
Log event matching, clock fires, DAG node execution, outbound sends and timeouts so automation runs are easier to trace. Co-authored-by: Cursor <cursoragent@cursor.com>
Give the automations page a Dify-style palette, drag-and-connect canvas, and typed inspector, and persist layout plus catalog/validate APIs for the editor. Co-authored-by: Cursor <cursoragent@cursor.com>
Keep stats and selection on one viewport and the graph editor on the next, snapping by scroll instead of replacing the list. Co-authored-by: Cursor <cursoragent@cursor.com>
Replace the blank-LLM multi-selects with searchable toggles, and connect edges by clicking an outlet then a valid target. Co-authored-by: Cursor <cursoragent@cursor.com>
Keep the 自动化 header outside the list/canvas scroll so it stays visible like other sticky page headers. Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
…on context Ready nodes now start as soon as their deps complete instead of waiting on a gather wave, and message-triggered tools inherit the current session so send_message can omit the target. Co-authored-by: Cursor <cursoragent@cursor.com>
…runtime Tool nodes were reading an empty context and reporting cognitive memory as disabled even when the bot had it enabled. Co-authored-by: Cursor <cursoragent@cursor.com>
Saving stayed on the canvas and skipped the list refresh while busy, so cards kept stale server state. Co-authored-by: Cursor <cursoragent@cursor.com>
Event automations no longer cool down by default, and concurrency, node/workflow timeouts, and blank LLM iterations are large enough for longer graphs. Co-authored-by: Cursor <cursoragent@cursor.com>
Event workflows with consume_ai_loop disabled no longer wait before the main AI, and new WebUI drafts default intercept and auto-send off. Co-authored-by: Cursor <cursoragent@cursor.com>
Let blank/agent/main LLM nodes declare extract_vars so the model writes named {{values}} through injected extract_* tools; WebUI inspector and canvas expose the same setting.
Co-authored-by: Cursor <cursoragent@cursor.com>
Use {{trigger.nickname}} instead of the QQ number, and resolve group card or QQ nickname on join/leave events so the greeting is not empty.
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: GPT-5 <noreply@openai.com>
|
You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard. |
Contributor
|
Important Review skippedToo many files! This PR contains 131 files, which is 31 over the limit of 100. To get a review, reduce the PR to 100 files or fewer by splitting it into smaller PRs or changing its base branch. Upgrade to a paid plan to raise the limit. This review couldn't start because sufficient usage credits or metered capacity aren't available. Add credits or update usage-based reviews in the billing tab, then retry. ⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Pro Plus Run ID: ⛔ Files ignored due to path filters (5)
📒 Files selected for processing (131)
You can disable this status message by setting the Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
…-blocking - poke: honor the core process_poke_message toggle at entry, skipping history writes, automation matching, and AI replies when disabled - group/private/wechat entries: stop matching automations when the corresponding should_process_* gate is off - consume_ai_loop now defaults to false across service, migration, short commands, and API serialization so matched messages reach the main AI unless a workflow explicitly opts in - document gate semantics and the new default; cover them with tests
- loop_max_iterations is now purely config-driven with a floor of 1 and no ceiling; the previous 25 clamp is gone from config parsing, service settings, runner execution, and validation - validate_automation/collect_automation_issues accept the configured cap so save-time checks match runtime; node max_iterations must be 1..configured - catalog endpoint exposes the configured value instead of a fixed 25; WebUI loop inputs drop the max=25 attribute - docs, changelog and config example updated; tests cover parsing, validation and runner behavior under larger caps
Co-authored-by: GPT-5 Codex <noreply@openai.com>
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.
本版本将「定时任务」整体升级为条件驱动的「自动化 / Automations」工作流：消息与时间事件命中后按 DAG 执行工具、模板与 LLM 节点，可选择接管本轮 AI；WebUI 配套 Dify 式画布编辑器，旧定时任务数据在启动时自动迁移。
src/Undefined/automations/：支持场景多选与 @ 专项消费，节点覆盖 tool / template / 三种 LLM，可自动派生 if-else 与 LLM 分支（分支选项即工具调用），循环最多 25 次。运行时只写data/automations.json；启动时若发现旧scheduled_tasks.json则自动转为新格式写入新文件，旧文件保留、不双写。AutomationService取代原调度服务，时间 job 只携带task_id；/api/v1/schedules与scheduler.*随之下线，对外只剩/api/v1/automations与automation.*的增删改查加启停共六个工具。consume_ai_loop=true时等工作流完成并拦截该入口的 AI 回复，否则后台执行、立刻放行主 AI。send_message只填message即发往当前会话；工具节点自动注入认知记忆、知识库、表情包与附件注册表，避免已启用能力被误报未启用。{{名称}}引用；三种 LLM 还支持extract_vars，把待抽取变量变成extract_<名称>工具由模型调用后写入；入退群事件会解析群名片 / QQ 昵称写入{{trigger.nickname}}，入群欢迎预设直接可用。POST /api/v1/automations/validate一次性返回全部问题；配置节[automations]支持热更新。default_cooldown_seconds = 0），并发与节点 / 工作流超时上限提高，空白 LLM 迭代上限放宽到 100；新建自动化默认不拦截主 AI、不自动发送终值；事件匹配、时间触发、节点执行、出站发送与超时均有详细日志。