Hacktoberfest 2026: the issues maintainers tagged for October, open and beginner-friendly. Browse Hacktoberfest issues

feat(plaita-ai): flow-coder 形态感知——作者期两问 + wake_service 信号进 skill

Open Beginner friendly
  • Is something wrong?
#348 0 comments 0 reactions 0 assignees View on GitHub

Maintainers usually reply within 1 day

Nobody has claimed this yet.

Assessment

Difficulty
2/5
Estimated time
1-3 hours
Newbie friendliness
84/100
Issue type
Documentation
Clarity
Clearly specified
Activity status
Active
Tech stack
python
Domain
developer-experience, documentation

Research direction

Update SKILL.md’s first “Understand the request” step and references/authoring-spec.md §7, following the existing flow-coder evals. First read those sections and the flow_compile(profile=) and flow_list_nodes guidance described in the issue; add the two shape questions, the inline/managed/managed:queue decision guidance, and a compile self-check example. Done when both documents cover the acceptance criteria and the existing evals pass without regression.

Written by the indexing model from the issue text.

Description

优先级 P2 · 依赖：无

基线 HEAD = 6f6fb862

Summary

AI 作者路径（flow-coder skill）对执行形态零感知，作者期与执行期的形态信息断层：

  • SKILL.md:288 仅一句运维向提醒（「worker 侧要注入业务节点，否则发布成功、调度必败」）——是部署提示不是形态感知；
  • 第 1 步「理解需求」无形态问题：AI 不知道先问「流程要不要挂起等人/定时」，写完才在执行闸撞墙（虽具名可修，但多一轮返工）；
  • 引擎侧作者期信号已落位：flow_compile(profile=) 附非致命 execution_requirements 报告、flow_list_nodes 行内 tier（inline/managed）+ suspends 字段（#331 后半 + 台阶化）——skill 侧未消费；
  • references/authoring-spec.md §7 发布链路无「先定形态」步骤。

影响

  • AI 写出的 flow 在错误的宿主上必败（如带 delay 的 flow 跑直跑），依赖执行期具名错误兜底——多一轮「报错→改」循环；
  • AI 无法在选节点阶段告知用户「这个方案需要托管面/需要配套服务」。

建议

  1. SKILL.md 第 1 步「理解需求」加形态两问：流程要不要挂起（等人/定时/外部消息）？要的话目标宿主是 console 托管还是嵌入直跑？——答案决定节点选择（挂起类节点只在托管方案出现）；
  2. authoring-spec.md §7 发布链路前插「先定形态」小节：目标环境（inline/managed/managed:queue 三行表）+ flow_compile(profile=...) 自检示例（非致命报告读法）；
  3. flow_list_nodes 消费：skill 指引 AI 选节点时读 tier 字段（managed 族标注「需托管」，delay/approval 族标注「需队列唤醒服务」）；
  4. MCP flow_compile docstring 的 profile 说明同步两级词汇（inline/managed，旧档名归一化兼容）——#331 作者期信号测试已覆盖，本单只改 skill 文本与指引。

验收条件：

  • SKILL.md 第 1 步含形态两问且有决策指引（挂起 → 托管）；
  • authoring-spec §7 含「先定形态」+ flow_compile(profile) 自检示例；
  • flow-coder evals 既有用例零回归（文本改动不破评测）；
  • 引擎侧 flow_compile/flow_list_nodes 零改动（信号面已落位，纯 skill 消费）。

边界

  • 不做自动宿主探测（profile 由调用方/用户给出）；
  • 不改 MCP 工具签名（既有参数已够用）。
Dominant language
Python
Stars
0
Forks
1
Avg merge
1h 55m
Merged PRs (30d)
13

Getting set up

This project ships no dev container, Dockerfile or contributing guide, so setting up is up to you: start from its README, and see our first-contribution guide for the general steps.

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from jeffkit/plaita

All issues in jeffkit/plaita

Similar issues

More Python issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.