Agent-Skills.md

Agent Skills: Long Task Management

Initialize or resume a long-running task session. Use when starting a complex multi-session task, resuming work from a previous session, or when the user mentions claude-progress.json or long-running work.

UncategorizedID: genomewalker/cc-soul/long-task

Repository

genomewalker/cc-soul
genomewalker
1

Install this agent skill to your local

pnpm dlx add-skill https://github.com/genomewalker/cc-soul/tree/HEAD/skills/long-task

Skill Files

Browse the full folder contents for long-task.

Download Skill

Loading file tree…

skills/long-task/SKILL.md

Skill Metadata

Name
long-task
Description
Initialize or resume a long-running task session. Use when starting a complex multi-session task, resuming work from a previous session, or when the user mentions claude-progress.json or long-running work.

Long Task Management

Mind-powered persistence for tasks that span multiple sessions.

Quick Commands

  • /long-task - Show active task or start new one
  • /long-task status - Show current task status
  • /long-task complete - Mark task as completed

Starting a Task

When starting a long-running task:

Use long_task_start with:
- task_id: unique identifier (e.g., "implement-feature-x")
- goal: clear description of what you're trying to achieve
- hard_checks: deterministic completion criteria (optional)
- soft_checks: semantic completion criteria (optional)

How It Works

  1. SessionStart: If active task exists, context auto-injects
  2. During work: Log significant events with long_task_event
  3. Stop hook: Automatically logs checkpoint and increments iteration
  4. Completion: Mark with [TASK_COMPLETE] or call long_task_complete

MCP Tools

| Tool | Purpose | |------|---------| | long_task_start | Start a new long-running task | | long_task_get | Get task details by ID | | long_task_active | Get active task for realm | | long_task_update | Update progress (completed_summary, work_items, blockers) | | long_task_complete | Mark task as completed | | long_task_event | Log event (decision, error, checkpoint) | | long_task_snapshot | Get synthesized context | | long_task_evaluate | Check completion criteria |

Process

If no arguments (show status or start)

  1. Check for active task with long_task_active
  2. If found: show snapshot with long_task_snapshot
  3. If not found: ask user for task goal and start new task

If "status" argument

  1. Get active task with long_task_active
  2. Show detailed snapshot with mode="debug"

If "complete" argument

  1. Get active task
  2. Ask for outcome summary
  3. Call long_task_complete

Example Usage

Start a task:

/long-task
> No active task. What's your goal?
> Implement the authentication system

Starting task: implement-auth
Goal: Implement the authentication system

Check status:

/long-task status
[LONG_TASK:implement-auth] Implement the authentication system
Iteration: 3 | Status: active
Done: Added login form, API endpoints
Pending: Add password reset; Add OAuth

Complete task:

/long-task complete
Outcome: Authentication system fully implemented with login, logout, and OAuth

Completed task: implement-auth

Completion Markers

The Stop hook detects these markers to auto-complete:

  • [TASK_COMPLETE]
  • [DONE]
  • <promise>COMPLETE</promise>

Multi-Agent Support

Tasks support lease-based coordination:

  • task_claim: Claim task with lease
  • task_heartbeat: Extend lease
  • Expired leases allow other agents to take over