Hand the current interactive Claude/Codex session off to an unattended agent-jdi background run — it quits this session and resumes it autonomously so the human can walk away. Trigger when the user says "hand this off to jdi", "justdoit", "let jdi finish this", "hand off to agent-jdi", "take it from here unattended", or otherwise wants the agent to keep working on its own after they leave.
Scanned 9/2/2026
Install to Claude Code
npx -y skills add tanghong123/claude-replay --skill jdi-handoff --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Jdi Handoff?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/tanghong123-jdi-handoff)More formats (shields.io, HTML) on the badges page.
---
name: jdi-handoff
description: Hand the current interactive Claude/Codex session off to an unattended agent-jdi background run — it quits this session and resumes it autonomously so the human can walk away. Trigger when the user says "hand this off to jdi", "justdoit", "let jdi finish this", "hand off to agent-jdi", "take it from here unattended", or otherwise wants the agent to keep working on its own after they leave.
---
# jdi-handoff
Hand the current session to `agent-jdi` so it continues **unattended** in the
background. This is the mirror of `agent-jdi takeover` (which hands a background
run back to a human).
## When to use
The user wants to stop babysitting the session and let it finish on its own —
"hand this off to jdi", "justdoit", "let jdi take it from here", etc.
## How
1. Figure out a short instruction for what remains — ask the user, or summarize
the current goal in one line.
2. Run, from **inside** this session (it finds the session automatically):
```
agent-jdi handoff <instruction>
```
This spawns a detached watcher and then quits this session. Once the session
exits, agent-jdi resumes it unattended and drives the work to completion,
retrying on recoverable failures.
For Codex, handoff preserves the current Codex permission context: `read-only`,
`workspace-write` with its exact network setting, or `danger-full-access`.
Capture happens before the watcher starts or the current session exits. If the
current turn's permission policy cannot be read safely, handoff aborts and
leaves the interactive session running. Retries and backlog drains reuse the
captured policy. No extra Skill flag is required. Claude permission behavior
is unchanged.
Do not ask for extra confirmation once the user has asked to hand off — just run
the command.
## Options
- `agent-jdi handoff --armed <instruction>` — arm the handoff **without** quitting;
the user presses `/exit` themselves when ready.
- `agent-jdi handoff --interval <secs> --max-attempts <n> <instruction>` — tune the
unattended retry loop.
## Getting it back
To pull the unattended run back to an interactive session later:
```
agent-jdi takeover
```
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!