Common Phase Instructions

Standardize workflow phase agent file operations and state tracking.

1|1|Updated Feb 1, 2026
One-click install
npx skills add https://github.com/zb-ss/claude-plugin-workflow --skill common-phase-instructions
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Common Phase Instructions
Source: https://github.com/zb-ss/claude-plugin-workflow/tree/main/skills/phases/common
Command: npx skills add https://github.com/zb-ss/claude-plugin-workflow --skill common-phase-instructions

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It standardizes how all workflow phase agents perform file operations, update workflow state, and reference codebase context so actions stay deterministic and the workflow doesn’t drift.

Core Features & Use Cases

  • Tool-safe file I/O: Enforces Claude Code native Read/Write/Edit tools for all file operations to prevent brittle or unsafe shell-based manipulation.
  • Correct home-path handling: Prohibits ~ expansion and requires absolute paths derived from the home directory, reducing path-related failures.
  • Consistent state updates: Requires a repeatable pattern for marking status in-progress/completed and logging errors directly into the workflow state file.
  • Context efficiency guidance: Directs agents to read only relevant sections and to avoid embedding full codebase context into subagent prompts.
  • Use-case: Keeps multi-step phase agents (e.g., build, review, test) aligned when they coordinate work across iterations and context compaction.

Quick Start

In a workflow phase prompt, instruct the agent to follow the Common Phase Instructions exactly for tool usage, state file updates, and context loading for the current task.

Frequently Asked Questions about Common Phase Instructions

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I keep workflow agents consistent and prevent state drift during multi-phase execution?

To keep workflow agents consistent, standardize their behavior by enforcing safe file I/O operations, requiring absolute paths, and mandating a single source-of-truth workflow state file for tracking status across iterations.

Why do my agent orchestration file operations fail when using shell commands and tilde paths?

Agent orchestration file operations fail when using shell commands and tilde paths because the standardized approach prohibits brittle shell manipulation and requires absolute paths derived from the home directory alongside native Read/Write/Edit tools.

What is the best way to manage codebase context loading for multi-step phase agents?

The best way to manage codebase context loading for multi-step phase agents is to apply context efficiency constraints, directing agents to read only relevant sections with offset and limit parameters rather than embedding full codebase context into subagent prompts.

How do I standardize workflow state updates across build, review, and test agents?

Standardize workflow state updates across build, review, and test agents by enforcing a repeatable pattern that marks status as in-progress or completed and logs errors directly into the single source-of-truth workflow state file.

Does agent orchestration work without context compaction issues during long iterations?

Agent orchestration does work without context compaction issues by enforcing context efficiency constraints, directing agents to read only relevant sections with offset and limit reads, and coordinating work across iterations without embedding full codebase context.

When do I need to enforce deterministic state update patterns for workflow automation?

You need to enforce deterministic state update patterns for workflow automation when running multi-phase agent execution, ensuring that reading codebase context files and updating workflow state remain predictable and the workflow doesn't drift across iterations.