跳到正文
cccskills
免费在 GitHub 公开

project-workflow

Runs multi step work the way a careful senior engineer would. Triage before code, a short PRD in prds/, task.md kept current, docs synced after, and lessons recorded when a mistake repeats. Use when a task spans several files or steps, changes architecture, needs a plan the user can review, or when the user says "write a PRD", "plan this", "track this in task.md", or "what did we learn".

中文简介正在准备中。当前显示原文说明。

查看安装方法

包含的文件(5)

  • SKILL.md4.4 KB
  • agents/openai.yaml91 B
  • assets/large-logo.png3.2 KB
  • assets/small-logo.svg1.2 KB
  • references/prd-template.md1.9 KB

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

Project workflow

One loop for any task bigger than a typo. Plan in a PRD, track in task.md, sync docs when done, write down what went wrong so it does not repeat.

Skip this for one file fixes. A typo, a copy change, a single obvious bug. Just fix those.

The loop

  1. Triage. Read the request and the code before touching anything.
  2. PRD. Write prds/<slug>.md from the template. Share it if the plan has a real choice in it.
  3. Track. Add the work to task.md under ## To Do. Move items as they land.
  4. Build. Small commits worth of change at a time. Verify each step.
  5. Sync docs. Hand off to the project-docs skill to update task.md, changelog.md, files.md.
  6. Learn. If the user corrected the same thing twice, add a line to prds/lessons.md.

Triage

Before writing code, answer these in a few lines:

  • What is missing, broken, or incomplete?
  • What are two or three likely causes or approaches?
  • Which one is most likely, and why?
  • What is still unclear enough to ask about?

If the last answer is not empty, ask. One question, not five. Then wait.

For bugs, name the root cause before proposing the fix. A fix without a root cause is a guess.

PRD

Create prds/<feature-or-problem-slug>.md before non trivial work. Use the layout in references/prd-template.md.

Required sections:

  • Problem (one paragraph)
  • Root cause (bugs only)
  • Proposed solution
  • Files to change
  • Edge cases
  • Verification steps
  • Task completion log

Metadata at the top, always in UTC:

Created: YYYY-MM-DD HH:mm UTC
Last Updated: YYYY-MM-DD HH:mm UTC
Status: Draft | In Progress | Done

Keep it short. A PRD that takes longer to write than the change did is too long. Two screens is plenty for most work.

PRDs are .md files and live in prds/. Old or sensitive PRDs move to prds/archive/.

Task tracking

task.md has three sections: ## To Do, ## In Progress, ## Completed.

  • New work goes under To Do with a link to its PRD.
  • Only one item In Progress at a time when possible.
  • An item moves to Completed after verification, not after the edit.
  • Completed entries carry a YYYY-MM-DD HH:mm UTC timestamp, the files touched, and the verification command or outcome.

Good entry:

- [x] 2026-09-15 18:40 UTC Fix duplicate heartbeat writes. prds/heartbeat-dedup.md. convex/stats.ts, src/hooks/usePageTracking.ts. Verified: npx convex dev clean, no OCC retries in dashboard for 10 min.

Bad entry:

- [x] fixed heartbeat

Docs sync

After each feature or fix, task.md, changelog.md, and files.md have to agree with the code. Load the project-docs skill for the rules. In short:

  • changelog.md follows Keep a Changelog. Dates come from git log --date=short, never from memory.
  • files.md gets new files added and stale descriptions fixed.
  • Claims about what shipped come from the diff, not from the plan.

Execution style

  • Keep the change set tight. The request sets the scope. Extra ideas go in the summary, not the diff. Load avoid-feature-creep if scope starts to drift.
  • One subagent per research question when the task is large enough to benefit. Do not parallelize edits to the same file.
  • Stop and re plan if the work stops making sense. Update the PRD, then continue.
  • Before calling it done, ask: would a staff engineer approve this diff as is?

Learning loop

prds/lessons.md is a flat list. One line per lesson, newest at the bottom:

- 2026-09-15 Do not add `returns` validators that duplicate the schema doc shape; use a shared validator constant instead.

Add a lesson when:

  • the user corrects the same pattern a second time
  • a verification step catches something the plan missed
  • a tool or API behaved differently from the docs

Read prds/lessons.md at the start of any session in that repo.

Never

  • Deploy, publish, commit, or push. The user does those.
  • Mark a task done without running the verification step written in the PRD.
  • Invent a date. Run git log or use the current UTC time.
  • Write a PRD for a one line change.

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

Keeps a change scoped to what was asked. Flags extra features, refactors, and 'while we are here' edits, and moves them to a suggestions list instead of the diff. Use when a request is small and the agent is about to expand it, when a PRD grows beyond its problem statement, or when the user says 'just do what I asked'.

中文简介正在准备中。当前显示原文说明。

waynesutton/builder-skills4062026年9月28日 更新

convex

免费

Routes Convex work to the right builder skill and covers the handful of rules that apply everywhere: validators on every function, withIndex over filter, idempotent mutations, internal functions for scheduling. Use when a task mentions Convex and no more specific skill matches, or as the first stop before picking one.

中文简介正在准备中。当前显示原文说明。

waynesutton/builder-skills4062026年9月28日 更新

Builds AI agents on the Convex agent component: threads, messages, tools that call queries and mutations, streaming, RAG with vector search, and workflows for multi step jobs. Use when adding a chat assistant, tool calling agent, or retrieval feature to a Convex app.

中文简介正在准备中。当前显示原文说明。

waynesutton/builder-skills4062026年9月28日 更新

Production patterns for Convex apps and the rules the @convex-dev/eslint-plugin enforces: validators, indexes, idempotent mutations, avoiding OCC conflicts, thin function wrappers, error handling. Use when reviewing Convex code, asking whether a pattern is right, setting up ESLint, or fixing write conflicts and slow queries.

中文简介正在准备中。当前显示原文说明。

waynesutton/builder-skills4062026年9月28日 更新

Creates reusable Convex components with defineComponent, a clean client wrapper, their own schema, and an npm publish setup. Use when extracting a feature into a package, building something for the Convex component directory, or when a component's functions are not showing up in the parent app.

中文简介正在准备中。当前显示原文说明。

waynesutton/builder-skills4062026年9月28日 更新

Schedules work in Convex: cron jobs in convex/crons.ts, one off scheduled functions with runAfter and runAt, batching large jobs, and cancelling or inspecting the queue. Use when something needs to run on a timer, later, or in the background, or when a cron is not firing.

中文简介正在准备中。当前显示原文说明。

waynesutton/builder-skills4062026年9月28日 更新

查看 waynesutton 的全部技能

报告此技能的问题