Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Manages context continuity across OpenClaw agent sessions. Defines a start, monitor, and wrap cycle: recover context from structured files at session start, track context window usage against defined thresholds during the session, and write a compact handoff file when closing. On first run, extracts and installs a companion buffer-optimizer skill for workspace auditing and setup.
Threshold-based alerts and a structured HANDOFF.md format prevent context loss without requiring manual note-taking between sessions.
When to use it
- Recovering project context at the start of a new agent session
- Getting warned before context window usage degrades task quality
- Handing off an in-progress task cleanly between sessions
- Reducing redundant file reads to lower prompt cache costs
- Auditing workspace boot payload and AGENTS.md structure
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: 'Continue where we left off on the API refactor'
- 1Read HANDOFF.md to identify stopping point and next steps
- 2Read MEMORY.md to load current priorities and project state
- 3Resume work from the documented next step without reading unrelated files
- 4Run session_status to check context window usage and apply intake rules
- 5At wrap, overwrite HANDOFF.md with current state and confirm saved items to owner
Session resumed with full context, progress persisted to HANDOFF.md for the next session