AGENT SKILL

ralplan

Consensus planning with Planner/Architect/Critic iterative loop. Activate when user says: ralplan, consensus plan, consensus planning.

License 许可
Apache-2.0
最近更新
2026-10-31
Stars
★ 1
Forks
0

01 前言 Front Matter

name
ralplan
description
Consensus planning with Planner/Architect/Critic iterative loop. Activate when user says: ralplan, consensus plan, consensus planning.
argument-hint
[--deliberate] <task description>

03 SKILL.md 正文

Ralplan (Consensus Planning)

Shorthand for /plan --consensus. Triggers iterative planning with Planner, Architect, and Critic agents until consensus is reached.

Flags

  • --deliberate: Forces deliberate mode for high-risk work. Adds pre-mortem (3 scenarios) and expanded test planning.

Interactive Hook Protocol

MANDATORY: Use vscode_askQuestions at every decision gate in this skill (when available). If vscode_askQuestions is NOT available (e.g., Copilot CLI), present numbered options in markdown and ask the user to respond with a number or freeform text.

When to Fire Hooks

Trigger Point Question Type
Options presentation Present viable options with pros/cons for user selection
Architect concerns Surface architectural trade-offs for user decision
Critic rejection Show rejection reasons, ask user for direction
Consensus reached Final plan approval before execution
--deliberate pre-mortem Present risk scenarios for user prioritization

Workflow

  1. @planner creates initial plan with RALPLAN-DR summary:
    • Principles (3-5)
    • Decision Drivers (top 3)
    • Viable Options (>=2) with pros/cons
  2. HOOK: Present options via vscode_askQuestions when options have significant trade-offs:
    header: "ralplan-options"
    question: "Planner identified [N] viable approaches. Select preferred direction:"
    options: [
      { label: "Option A: [name]", description: "Pros: X, Y. Cons: Z" },
      { label: "Option B: [name]", description: "Pros: A, B. Cons: C" },
      { label: "Let agents decide based on technical merit", recommended: true }
    ]
    allowFreeformInput: true
    
  3. @architect reviews for architectural soundness
    • If architect raises concerns: HOOK — present trade-offs for user decision
  4. @critic validates quality and testability
    • If critic rejects: HOOK — show issues and ask for direction:
    header: "ralplan-critic-feedback"
    question: "Critic identified [N] issues: [summary]. How to proceed?"
    options: [
      { label: "Address all issues", recommended: true },
      { label: "Address critical issues only, skip minor" },
      { label: "Override — I accept the trade-offs" },
      { label: "Restart with different constraints" }
    ]
    
  5. Loop until critic approves (max 5 iterations)
  6. Final plan includes ADR (Decision, Drivers, Alternatives, Why chosen, Consequences)
  7. HOOK: Final approval via vscode_askQuestions:
    header: "ralplan-approval"
    question: "Consensus reached after [N] iterations. Execute?"
    options: [
      { label: "Execute with team (parallel)", recommended: true },
      { label: "Execute with ralph (sequential + verification)" },
      { label: "Execute with omg-autopilot (full pipeline)" },
      { label: "Save plan — don't execute yet" }
    ]
    

Deliberate Mode Hooks

When --deliberate is active, add:

  • HOOK: Pre-mortem scenarios — present 3 risk scenarios, ask user to rank priority
  • HOOK: Test strategy — present expanded test plan for user approval

Pre-Execution Gate

Vague execution requests (e.g., "ralph improve the app") are redirected through ralplan first to ensure explicit scope, testable criteria, and multi-agent consensus.

Passes gate (specific enough): prompts with file paths, function names, issue numbers, numbered steps, or acceptance criteria.

Gated (needs scoping): prompts with only vague descriptions and no concrete anchors.

After Approval

  • Execute via /team (parallel agents, recommended) or /ralph (sequential with verification)

04 同仓库更多 Skills