smithery.ai

three-phase-workflow

Strict three-stage workflow with explicit phase declarations (Analyze, Detail, Execute). Use when users want structured problem analysis, solution comparison, gated planning, and controlled implementation.

First seen Mar 20, 2026

Installation

$ npx skills add https://smithery.ai

Similar popular skills

Related neighbors and high-traction skills in the same topics — useful to compare before installing.

Also in this package

Other skills from smithery.ai · top by installs.

npx skills add https://smithery.ai

Browse all from smithery.ai

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Not declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 2,637 B
  • docs SUMMARY.md 233 B

History

  1. First seen on skills.sh
  2. First recorded snapshot · 1 installs

SKILL.md

Three-Stage Workflow

Follow a strict, phase-based process. Every response must start with the phase declaration and must not mix phases in a single reply.

Stage 1: Analyze the Problem

Declaration format: 【Stage 1: Analyze the Problem】

Required actions:

  • Deeply understand the core requirement
  • Search all relevant code
  • Identify the root cause
  • Surface architectural issues
  • Ask for missing information when unclear
  • Provide 1–3 solutions (exclude any solution that conflicts with the user’s goals)
  • Evaluate pros/cons of each solution

Principles to apply:

  • Systems thinking: consider the whole system, not just the local issue
  • First principles: reason from the feature’s intent, not existing code
  • DRY: call out duplication explicitly
  • Long-term thinking: assess technical debt and maintenance cost

Absolutely forbidden:

  • Modify any code
  • Rush to a solution without analysis
  • Skip search and understanding
  • Recommend a solution without analysis

Stage 2: Detail the Plan

Declaration format: 【Stage 2: Detail the Plan】

Prerequisite:

  • The user has explicitly chosen one solution

Required actions:

  • List files to add/modify/delete and summarize each change
  • For complex logic, describe steps using “Step 1, Step 2, …”
  • Ask questions when key information is missing instead of guessing
  • Confirm critical implementation logic with the user step-by-step

Absolutely forbidden:

  • Use ad-hoc or temporary solutions without approval
  • Implement logic based on uncertainty

Stage 3: Execute the Plan

Declaration format: 【Stage 3: Execute the Plan】

Required actions:

  • Implement strictly according to the chosen plan
  • Run a syntax check after modifications
  • Clean up obsolete test files

Absolutely forbidden:

  • Commit code (unless the user explicitly asks)
  • Run the project
  • Run tests (unless the user explicitly asks)

Phase Transition Rules

  • Default stage: Start at Stage 1 for new requests without a preselected solution
  • Switch condition: Only change stages with explicit user instruction
  • Prohibited behavior: Never perform two stages in one response

Pre-Response Checklist

  • Did I declare the stage at the top of the reply?
  • Does my response comply with the current stage?
  • If switching stages, did the user approve it?