Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
Refactor: split monolithic SKILL.md into modular component files; keep behavior/output contract unchanged.
What This Skill Does
Runs a structured end-of-session workflow triggered by phrases like 'wrap up' or '/wrap-up'. Executes four ordered phases: ship pending changes, consolidate agent memory, apply self-improvements, and produce a report with safety gates between each step.
Replaces a manual end-of-session checklist by running ship state, memory consolidation, and report generation in one command with safety gates that prevent skipping steps or losing work.
When to use it
- Committing and pushing code at the end of a work session
- Generating a wrap-up summary after a long multi-step agent task
- Consolidating session decisions and notes into persistent memory
- Applying queued self-improvement rules before closing an agent loop
- Producing a machine-readable JSON artifact of completed session work
Example Workflow
Here's how your AI assistant might use this skill in practice.
User says: wrap up
- 1Reads design principles and action gates from component 01
- 2Executes Ship State phase, staging and committing pending changes
- 3Runs Memory phase, consolidating session notes into persistent memory
- 4Applies self-improvements and processes the publish queue
- 5Assembles the final inline report with both human-readable and JSON artifacts
A consolidated report with two artifacts: a human-readable wrap report and a structured JSON summary of the closed session