skills.rest
tmustier
tmustierCOMMUNITY
Explore all skills by this author →
Website𝕏@thomasmustierGitHub

long-task-harness

Preserve session progress and feature state across long-running agent tasks.

3|Updated Dec 1, 2025
Search Tags:#context management#progress tracking#session continuity#multi-session#feature management
One-click install
npx skills add https://github.com/tmustier/long-task-harness --skill long-task-harness
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: long-task-harness
Source: https://github.com/tmustier/long-task-harness/tree/main/.
Command: npx skills add https://github.com/tmustier/long-task-harness --skill long-task-harness

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill solves the "shift change" problem where AI agents lose context between sessions, ensuring your long-running projects stay on track without manual catch-up work.

Core Features & Use Cases

  • Progress Documentation: Automatically maintain structured session logs with goals, accomplishments, and next steps.
  • Feature Tracking: Maintain a comprehensive feature checklist with pass/fail status and history tracking.
  • Context-Efficient Workflows: Load only relevant history with 78% context reduction.
  • Session Handoff: Provide smooth transitions between AI agents with bidirectional linking and decision tracking.

Quick Start

When starting a complex project that will span multiple sessions, invoke the long-task-harness skill to initialize tracking files and configure automatic invocation for your agent.

Quick Start

I'm starting a new multi-session project. Please initialize the long-task-harness tracking system for this project.

Frequently Asked Questions about long-task-harness

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

FAQPage Schema
How do I prevent losing context between AI agent sessions?▼

Context loss between sessions occurs when agents restart without access to prior work. Long-task-harness preserves progress, feature state, and session history in structured files (.long-task-harness), enabling agents to resume projects seamlessly with 78% context reduction and bidirectional session linking.

What's the best way to track progress across multi-session development projects?▼

Maintain structured session logs with progress.md documenting goals, accomplishments, and next steps for each session. Long-task-harness automates this tracking, storing session metadata and feature checklists with pass/fail status so agents can load only relevant history on resume.

Can I use long-task-harness with git-based workflows?▼

Yes, long-task-harness integrates with git as a dependency. It enforces session hooks and pre-commit reminders to capture handoff context, storing .long-task-harness data (progress.md, features.json, init.sh) alongside your repository for version control and continuity.

How do I initialize context tracking for a new multi-session project?▼

Invoke long-task-harness initialization to set up tracking files and configure automatic invocation for your agent. It generates the .long-task-harness directory with progress documentation, feature checklist, and session handoff scripts for smooth agent transitions.

What happens if an AI agent doesn't check prior session history?▼

Without context loading, agents restart without decision history or feature status, duplicating work and losing continuity. Long-task-harness provides context-efficient scripts to load recent sessions and enforce pre-commit reminders, ensuring agents access prior context automatically.

Related Skills

Explore More Software Engineering Skills→
oeftimieoeftimie
19

harness-init

Initialize a persistent agent harness with git identity and project scaffolding.

Community
Advanced
Nate-VishNate-Vish
7

compact

Archive old agent sessions and refresh Context.md with summaries.

Community
Advanced
tygwantygwan
1

git-workflow

Enforce GitHub Flow, Conventional Commits, and PR processes across repositories.

Community
Intermediate
matthew-h-cromermatthew-h-cromer

commit-conventions

Provide guidelines for writing structured git commit messages.

Community
Basic