vibe-action 0.0.4

Command router — execute shell commands and LLM prompts via simple YAML actions.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35

name: review
about: Critically analyze code for bugs and flaws using a powerful model
clipboard: true
notify: true
args:
  - name: query
    short: 'q'
    expect: string
    help: The code block or function to review
    default: '{system_clipboard}'
  - name: to
    short: 'l'
    expect: string
    help: Target language (e.g., Russian, English, Chinese)
    default: Russian
actions:
  # Analyze code for bugs and generate technical recommendations in the target language
  - tag: tag_review
    run: llm_large
    expect: string
    action: |
      [Task]
      Analyze the provided code for logic flaws, concurrency bugs, memory leaks, and performance bottlenecks.

      [Rules]
      - Write strictly in the language specified in [User]. Keep feedback concise, blunt, and highly technical.
      - If the code has no flaws, output exactly one word: PERFECT.
      - Format each issue as a clear block: standard bullet point for the description, followed by a new line with a prefix "  Fix: " and the exact 1-line code patch.
      - Do NOT write explanation inside the fix line, provide ONLY the clean code patch.

      [User]
      {to}

      [Code to Review]
      {query}