# Reboot

> Distill the current task and state into a clean, self-contained reprompt to paste into a fresh session, so a stale or bloated context can be cleared without losing the thread. Triggers "/reboot", "reboot this session", "this session is stale", "stale session", "clear and reprompt", "reprompt me", "fresh start", "this session's getting old". Do NOT use for a cross-terminal morning digest (that's m…

- **Type:** Skill
- **Install:** `agentstack add skill-ao92265-claude-code-playbook-reboot`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [ao92265](https://agentstack.voostack.com/s/ao92265)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [ao92265](https://github.com/ao92265)
- **Source:** https://github.com/ao92265/claude-code-playbook/tree/main/skills/reboot

## Install

```sh
agentstack add skill-ao92265-claude-code-playbook-reboot
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# reboot

The session is old or bloated and should be cleared. Your job: hand the user a clean, self-contained prompt they can paste into a **fresh** session and keep going — no thread lost.

You can't run `/clear` yourself. You produce the reprompt; the user clears and pastes.

## Sources

- **Primary: the current conversation** you already hold. That's the real source of the goal and state.
- **Cross-check / fill gaps:** the freshest handoff in `~/.claude/handoffs/` — pick the one matching the current working directory, else the most recently modified — and its `.compact.md` sibling if present. These are the same sources `sessionstart-handoff.sh` re-injects into a new session.

## Distill into the reprompt

Capture only what a fresh Claude needs:

- **Goal** — what "done" looks like, in one or two lines.
- **State** — what's already done / decided.
- **Next step** — the immediate thing to do.
- **Constraints** — hard rules or decisions already locked in.
- **Key paths** — files/dirs in play.

Drop dead-ends, resolved tangents, and tool noise.

## Output

Two short lines first: why the session looks stale, and what to do — `/clear` (or open a new terminal), then paste the block below.

Then **one fenced code block** = the paste-ready reprompt. Write it as a direct instruction to a fresh Claude. It must be fully self-contained — assume the reader saw none of this session. No "as discussed", no dangling pronouns.

## Honesty

Don't invent state. If branch, dirty files, or anything else is uncertain, write "verify before acting" instead of asserting it — same disclaimer the handoff itself uses. A wrong "current state" is worse than an admitted gap.

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [ao92265](https://github.com/ao92265)
- **Source:** [ao92265/claude-code-playbook](https://github.com/ao92265/claude-code-playbook)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-ao92265-claude-code-playbook-reboot
- Seller: https://agentstack.voostack.com/s/ao92265
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.
