Can I manually resume a workflow?

Yes, you can use the 'find' command to locate a workflow by issue number and the 'get' command to retrieve the full context for resumption.

Where is the workflow data stored?

The state is stored locally in a git-ignored SQLite database located at .claude/execution-state.db.

Which phases are supported in the workflow?

The system supports five standard phases: setup, research, implement, review, and finalize.

What is the purpose of the checkpoint system?

It allows AI agents to save their progress so they can resume work if the conversation context is compacted or the session is interrupted.

Workflow Checkpoint Manager

Name: Workflow Checkpoint Manager
Author: joeczar

byjoeczar

Productivity & Workflow

Persists and manages AI agent session states to enable seamless task resumption after context compaction or interruptions.

About

The Checkpoint Workflow skill provides a structured framework for Claude to track its progress during complex coding tasks. By maintaining a local database of current phases, logged actions, and git commits, it ensures that AI agents can resume work precisely where they left off if context limits are reached or a session is restarted. This skill is essential for maintaining continuity in long-running development cycles, ensuring that research, implementation, and review phases are preserved even across different terminal sessions.

Key Features

0 GitHub stars
Detailed action logging for historical context and auditability
Persistent state tracking across multiple AI coding sessions
Atomic git commit tracking with unique SHA verification
Phase-based workflow management from research to finalization
Automatic resume detection to prevent duplicate work or lost context

Use Cases

Maintaining continuity when switching between different agents or task phases
Tracking task progress for long-running issues that span several days
Resuming complex feature implementation after hitting context window limits

About

Key Features

0 GitHub stars
Detailed action logging for historical context and auditability
Persistent state tracking across multiple AI coding sessions
Atomic git commit tracking with unique SHA verification
Phase-based workflow management from research to finalization
Automatic resume detection to prevent duplicate work or lost context

Use Cases

Maintaining continuity when switching between different agents or task phases
Tracking task progress for long-running issues that span several days
Resuming complex feature implementation after hitting context window limits