From aiagentminder
Session checkpoint — capture state for clean resume
How this skill is triggered — by the user, by Claude, or both
Slash command
/aiagentminder:handoffThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Capture current state so work can be resumed cleanly. Run this before ending a session.
Capture current state so work can be resumed cleanly. Run this before ending a session.
Review what happened this session:
**Status:** in-progress. If so, note which issue was active and the done/in-progress/todo/blocked counts.If any of these happened this session, add an ADR entry:
## Known Debt section)Use the format already established in DECISIONS.md. Always include alternatives considered and the tradeoff accepted — a decision without alternatives is an assertion, not a record.
Locate the project's auto-memory file at:
~/.claude/projects/[project-path-hash]/memory/MEMORY.md
Append or update a "Next Session" section:
## Next Session
- [Most important thing to know picking up next — be specific, not vague]
- [Second priority or open question]
- [Third item only if genuinely needed]
Keep each bullet to one sentence. Do not exceed 3 items. Specificity matters: "Continue API work" is not useful; "Implement POST /users/:id — GET is done, need POST with request validation" is.
git add DECISIONS.md
git commit -m "handoff: session checkpoint [today's date]"
Only stage files that were modified. Do NOT modify SPRINT.md during handoff — sprint state is updated during sprint execution.
Session handoff complete.
This session:
- [what was accomplished — specific files/features/fixes]
State of things:
- [what's working, what's not, what's in progress]
[if sprint active:]
Sprint [n]: [done]/[total] issues done[, [blocked] blocked]
Currently on: S[n]-[seq] — [issue title]
Next session should:
1. [specific first action]
2. [specific second action]
Blockers for human:
- [anything requiring human action, or "None"]
Guides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.
npx claudepluginhub lwalden/aiagentminder --plugin aiagentminder