checkpoint

Persist and resume dotagent workflow state across phases after interruptions.

Updated Dec 1, 2025
One-click install
npx skills add https://github.com/Clouder0/dotagent --skill checkpoint-clouder0
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: checkpoint
Source: https://github.com/Clouder0/dotagent/tree/main/.claude/skills/checkpoint
Command: npx skills add https://github.com/Clouder0/dotagent --skill checkpoint-clouder0

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Robust workflow checkpointing and resume capabilities to recover from interruptions.

Core Features & Use Cases

  • Phase checkpointing: Save and resume phases without losing progress.
  • State integrity checks: Validate references and outputs before resuming.
  • Use Case: After a crash, resume from the last valid phase and continue execution.

Quick Start

Start a workflow and, if interrupted, run Claude with resume to continue from the last checkpoint.

Frequently Asked Questions about checkpoint

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

FAQPage Schema
How do I resume a workflow after it crashes or gets interrupted?

Resuming workflows after interruptions uses checkpointing to save phase state automatically. When a workflow stops due to crash, timeout, or explicit interruption, run Claude with the resume command to pick up from the last valid checkpoint and continue execution without losing progress.

What happens to my workflow state when it gets interrupted mid-execution?

Workflow state is persisted to disk under memory/state and memory/reports using atomic writes, so interruptions don't cause data loss. The checkpoint system validates state integrity and coordinates recovery across all workflow phases—from REQUIREMENTS through REFLECTION—enabling safe continuation from the last verified point.

Can I replay a workflow from a specific checkpoint?

Yes, checkpointing supports explicit replays by restoring from saved phase state. You can resume execution from any archived checkpoint, allowing you to re-run phases or experiment with different paths without starting the entire workflow over.

Does checkpointing work with long-running dotagent workflows?

Checkpointing applies to long-running dotagent workflows across all execution phases. It handles interruptions, crashes, and timeouts by maintaining per-phase and per-task recovery coordination, making it suitable for workflows that span extended periods or multiple processing stages.

How does the system validate checkpoint integrity before resuming?

The checkpoint system enforces file-based persistence, atomic write protocols, and state integrity validation that checks references and outputs before resuming. This ensures corrupted or incomplete checkpoints are detected, preventing invalid state from propagating into resumed execution.

Related Skills