Write or update the handoff file so the next session (or another person/agent) can resume without re-discovering everything. Use when pausing or ending work, or when asked to produce a handoff.
Scanned 9/2/2026
Install to Claude Code
Skills Directorynpx -y skills add HaipingShi/coderail --skill handoff --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Handoff?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/haipingshi-handoff)More formats (shields.io, HTML) on the badges page.
---
name: handoff
description: Write or update the handoff file so the next session (or another person/agent) can resume without re-discovering everything. Use when pausing or ending work, or when asked to produce a handoff.
---
# handoff
Use only when H1/H2/H3 trigger fires.
## Before writing handoff
1. Run or read `/coderail:inspect` so the handoff reflects current state.
2. Confirm the active Coordinate G/T/S/V/X/P.
3. Confirm verification gaps are either resolved or explicitly marked blocked.
4. Write a handoff trace event if this is H1/H2/H3.
## Handoff content
Keep it short:
- Current North Star / current task
- Coordinate Summary, including Full Rail or Light Rail
- Recent completed slice
- Next executable step
- Stop conditions
- Generated and ignored artifacts
- Recovery commands
- Auto-commit action
Move long history to TRACE_INDEX, RUNLOG, DECISIONS, or archive. Do not paste full logs, full diffs, or TRACELOG contents.
Handoff is agent-to-agent continuity, not the Owner Receipt. Keep product
delivery facts in `docs/DELIVERIES.jsonl` and owner wording in `owner-summary`.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!
Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...
