Receber por email
← Voltar para o catálogo

code

Implement ONE pending Phase from subtask plan. Only write code, do not mark checkboxes or run tests. Called by autoworker:dispatch. Ends by calling autoworker:checkpoint.

16estrelas
Atualizado há 7 meses

Ver no GitHub ↗ Licença: MIT

Como adicionar

/plugin marketplace add phj128/autoworker

O comando exato pode variar conforme o repositório. Confira o README no GitHub.

Para o autor da skill

Cole no README do seu repo

Mostra que sua skill está catalogada na Skillteca, gera backlink e tráfego rastreável.

[![Skillteca](https://www.skillteca.com.br/api/badge/code/svg)](https://www.skillteca.com.br/skills/code)

Skills relacionadas

Ver todas de Desenvolvimento →

Alerta por categoria

Receba novas skills de Desenvolvimento toda segunda

1 email curto, só com as skills novas de Desenvolvimento. 4 minutos de leitura, sem spam, cancela com 1 clique.

Você confirma o email no primeiro envio. Sem spam. Cancela com 1 clique.

Ler descrição completa

autoworker:code — Implement One Phase

Called by autoworker:dispatch. Does one thing: implement the first incomplete Phase from the subtask plan.

Execution Flow

1. Read Subtask

Glob `subtask_*.md` (exclude subtask_template.md) →
  0 found → stop, prompt to create subtask
  1 found → use directly (backward compatible)
  multiple → grep `status:` to filter:
    - Files without status field treated as active (backward compatible)
    - Exactly 1 active → use it
    - 0 active → list all files + status, prompt user to choose
    - >1 active → report anomaly
→ Read → extract:
- Plan section Phase list (with checkboxes)
- Find the first incomplete Phase (`- [ ]` prefixed Steps)

If subtask doesn't exist or plan is empty → stop, prompt to complete autoworker:subtask-plan first. If all Phases are complete → stop, prompt that autoworker:dispatch should route to autoworker:test.

2. Implement Code

Read all Steps in the Phase, implement one by one:

  1. Read Step description and target file
  2. Read target file (if modifying)
  3. Edit/Write to implement the code change
  4. After completing the Step, continue to next Step

Executing commands: Steps may involve not just writing code but also executing commands (starting training, running scripts, etc.). Long-running commands use run_in_background=true — execute autonomously, do not ask the user to run manually. After starting background tasks like training, continue to the next step without waiting for completion.

Forbidden:

  • Writing code from memory without reading subtask
  • Skipping Steps (unless subtask explicitly marks as skippable)
  • Checking off boxes yourself (autoworker:checkpoint's responsibility)
  • Running tests yourself (autoworker:test's responsibility)
  • Making changes beyond the current Phase's scope
  • Asking user to manually execute commands that can be done autonomously

3. Output Summary

Phase X code implementation complete:
- Step X.1: <brief change description>
- Step X.2: <brief change description>
→ Invoking autoworker:checkpoint

4. Chain: Immediately Invoke autoworker:checkpoint

After outputting the summary, immediately invoke autoworker:checkpoint. Do not wait for user instructions, do nothing else.

Key Constraints

  • Only one Phase: Complete it and hand off to checkpoint — don't determine how many Phases remain
  • Only write code: Don't check boxes, don't run tests, don't make routing decisions
  • Self-debug when hitting issues: 2 consecutive failures with same approach → enter diagnostic mode (make assumptions explicit → minimum observable unit → build diagnostic tool)
  • Don't return to user mid-way: Unless hitting a genuine blocker requiring human decision

Important Notes

  • Phase granularity: A Phase typically contains 2-5 Steps, each Step is a specific file change
  • Files in current Phase: If a Step says "modify file X", must Read the file before Edit
  • Chaining is mandatory: Must invoke autoworker:checkpoint after completion, cannot skip or manually substitute
Compartilhar X LinkedIn