From john-skills
Saves and restores session state for continuity across context compressions. Creates checkpoints with task status, decisions, blockers, and next steps, and searches past checkpoints by semantic similarity.
How this skill is triggered — by the user, by Claude, or both
Slash command
/john-skills:memory-checkpointThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
**If user provides a query** (e.g., `/memory-checkpoint authentication bug`):
If user provides a query (e.g., /memory-checkpoint authentication bug):
→ Search checkpoints using scripts/search.py "<query>" --full --project-dir "$PWD"
If no query (e.g., /memory-checkpoint):
→ Create checkpoint using format from references/checkpoint-format.md, pipe to scripts/checkpoint.py --project-dir "$PWD"
Generate content with these sections, then pipe to script:
echo "CHECKPOINT_CONTENT" | python3 scripts/checkpoint.py --project-dir "$PWD"
python3 scripts/search.py "query" --full --project-dir "$PWD"
Returns checkpoints ranked by semantic similarity. --full includes content of top result.
./scripts/setup.sh
Creates .venv and installs fastembed. Scripts auto-detect the venv when present.
npx claudepluginhub jbdamask/john-claude-skills --plugin john-skillsManages session context across long development sessions using tiered summarization and checkpoints. Enables seamless resume after breaks.
Saves and resumes state for multi-phase commands like /debug, /epic, /feature, /implement, and /plan, enabling work to survive session interruptions.
Manages Claude Code session lifecycle: load project context, save checkpoints, and reflect on task progress with validation evidence. Use when starting, resuming, or concluding a session.