playbooks

home / skills / nickcrew / claude-cortex / session-management

session-management skill

/skills/session-management

This skill helps you manage session lifecycles by loading, saving, and reflecting on progress with structured checkpoints and summaries.

npx playbooks add skill nickcrew/claude-cortex --skill session-management

Review the files below or copy the command above to add this skill to your agents.

Files (4)
SKILL.md
1.1 KB
---
name: session-management
description: Use when loading, saving, or reflecting on session context - provides structured workflows for session lifecycle management and reflection outputs.
---

# Session Management

## Overview
Manage session lifecycle: load context, save checkpoints, and reflect on progress with consistent reporting.

## When to Use
- Loading project context or checkpoints
- Saving session context or learnings
- Reflecting on task completion and validation

Avoid when:
- The task is not session-related

## Quick Reference

| Task | Load reference |
| --- | --- |
| Load session | `skills/session-management/references/load.md` |
| Save session | `skills/session-management/references/save.md` |
| Reflect on session | `skills/session-management/references/reflect.md` |

## Workflow
1. Select the session operation.
2. Load the matching reference.
3. Execute the operation and capture outputs.
4. Provide a summary and next steps.

## Output
- Session status summary
- Saved context or reflection notes

## Common Mistakes
- Saving without capturing key changes
- Reflecting without validation evidence

Overview

This skill provides structured workflows to manage a session lifecycle: loading context, saving checkpoints, and producing reflective summaries. It standardizes how session data and learnings are captured so teams and agents can resume work consistently. Use it to maintain traceable session state and to generate concise next-step recommendations.

How this skill works

The skill lets you select an operation (load, save, reflect), loads the matching reference or checkpoint, executes the operation while capturing relevant outputs, and produces a summary with suggested next steps. Reflection runs include validation prompts to ground conclusions in evidence and surface gaps. Saved checkpoints include metadata to make resumption deterministic and auditable.

When to use it

  • Loading a project or agent context before resuming work
  • Saving important session state, changes, or discoveries as checkpoints
  • Running end-of-session reflections to summarize progress and blockers
  • Handing off work between team members or agents
  • Validating that completed tasks meet acceptance criteria before closing

Best practices

  • Capture key changes and rationale whenever you save a session checkpoint
  • Include validation evidence in reflections to avoid unsupported conclusions
  • Use clear, consistent metadata (timestamps, author, version) on saves
  • Keep reflections concise: outcomes, confidence, blockers, and next steps
  • Automate loading/saving where possible to reduce human error

Example use cases

  • Developer resumes a paused sprint task by loading the latest session context and checkpoint
  • Agent saves a learned configuration and reasoning trace after a multi-step run
  • Team lead runs a reflection at sprint end to summarize accomplishments and open issues
  • Automation pipeline checkpoints state before a risky migration and logs the outcome
  • Onboarding: provide a reproducible session snapshot for a new team member

FAQ

What does a reflection output include?

A concise summary of session status, validation evidence, learned items, confidence levels, and clear next steps.

How do I ensure saved checkpoints are usable?

Include structured metadata (timestamp, author/agent, version), the minimal reproducible context, and notes on dependencies to restore state.