How do I install the Claude Session Handoff skill?

Run npx skillmds@latest add batteryshark/claude-session-handoff in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Claude Session Handoff safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Claude Session Handoff?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Claude Session Handoff free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Claude Session Handoff?

batteryshark (@batteryshark) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Claude Session Handoff by batteryshark

Name: Claude Session Handoff
Author: batteryshark

Claude Session Handoff

Build a redacted, evidence-backed continuation brief from a Claude Code JSONL transcript. Use when resuming an interrupted agent task, handing work to another agent, or reconstructing objectives, edits, validation, and remaining work.

batteryshark 32ecfd6 5 files · 34.5 KB Updated Sep 17, 2026

File contents

Build a Claude Session Handoff

Identify the Claude Code JSONL transcript, the associated project root, the intended receiving agent, and whether the brief may contain private source or conversation material.

Read references/handoff-evidence.md, then run:

bin/claude-session-handoff /path/to/session.jsonl \
  --root /path/to/project \
  --redact-session-ids \
  --output /private/path/handoff.md

Keep default redaction enabled. Treat --no-redact as an explicit sensitive-data decision, not a convenience flag. Raw tool results are excluded unless --include-tool-results is requested because they may be large or secret-bearing.
Review the generated objectives, current-state signals, changed-file evidence, commands/results, MCP calls, task events, parse errors, and omitted-item counts against the transcript and current repository state.
Add a short verified continuation section stating what is actually complete, what remains, which assumptions are unresolved, and the first safe next action. Do not promote transcript claims to current truth without checking the workspace.
Deliver the brief through an approved private channel and retain the source line numbers needed for traceability.

Use agents/handoff-reviewer.md for an independent evidence and privacy pass.

Frequently asked questions