auto-resume

Reconstruct project context and execution state from git history and state files.

37|7|Updated Mar 31, 2026

Search Tags:#automation #git #workflow #context #recovery #state

Ready to run

npx skills add https://github.com/appautomaton/mlx-speech --skill auto-resume

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: auto-resume
Source: https://github.com/appautomaton/mlx-speech/tree/main/.claude/skills/auto-resume
Command: npx skills add https://github.com/appautomaton/mlx-speech --skill auto-resume

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires node, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill solves the problem of lost context and fragmented progress in long-running AI-assisted projects by rebuilding the state from durable artifacts rather than relying on unreliable memory.

Core Features & Use Cases

Durable State Recovery: Automatically reconstructs the active change, current stage, and pending tasks from project-specific state files.
Execution Ledger Reconciliation: Matches git history and file system status against the project plan to identify exactly where work was interrupted.
Use Case: If an AI session is terminated mid-project, this Skill allows the agent to immediately identify the last completed slice, surface pending engineering reviews, and determine the precise next step without manual re-briefing.

Quick Start

Run the auto-resume skill to rebuild the project context and identify the next required action from the current state.

Frequently Asked Questions about auto-resume

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

FAQPage Schema

How do I restore project context and resume interrupted work after an AI session is terminated?▼

To restore project context, this Skill automatically reconstructs the active change, current stage, and pending tasks from durable artifacts and git history. It matches file system status against the canonical plan to identify exactly where work was interrupted.

Do I need git CLI tools and a .agent directory to rebuild execution state?▼

Yes, rebuilding execution state requires access to the .agent directory and standard git CLI tools. These dependencies are necessary to verify the integrity of the execution ledger and reconcile current progress against the project plan.

What is the best way to recover pending engineering reviews from a fragmented development workflow?▼

The best way to recover pending reviews is through execution ledger reconciliation. This process matches git history against the project plan to surface pending tasks and determine the precise next step without manual re-briefing.

Can I use Node to automatically reconstruct the last completed slice of a multi-stage development workflow?▼

Yes, this Skill uses Node to automatically reconstruct the last completed slice. It reconciles durable state files to identify completed work and facilitate seamless session re-entry for multi-stage development workflows.

Why does an AI agent lose context in long-running projects and how do I prevent manual re-briefing?▼

AI agents lose context in long-running projects due to unreliable memory. This Skill prevents manual re-briefing by rebuilding state from durable artifacts, allowing the agent to immediately identify the last completed slice and pending tasks.

When should I not rely on durable artifacts for state recovery in development workflows?▼

You should not rely on durable artifacts for state recovery if your project lacks a canonical plan or git history. Without these elements, the Skill cannot reconcile the execution ledger or verify the integrity of your progress.