AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified MIT Self-run

Handoff

skill-jesse-merhi-skills-handoff · by jesse-merhi

Compact the current conversation into a handoff document for another agent to pick up. Prefer a fresh session/thread for standalone, parallel, background, or aside work; fork only when the next agent truly needs the raw current conversation history.

No reviews yet
0 installs
15 views
0.0% view→install

Install

$ agentstack add skill-jesse-merhi-skills-handoff

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-jesse-merhi-skills-handoff)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
26d ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Handoff? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Handoff

Write a handoff document summarizing the current conversation so a fresh agent can continue the work. Save it to the temporary directory of the user's OS, not the current workspace.

Include a "suggested skills" section in the document with the skills the next agent should invoke.

If the user passed arguments, treat them as a description of what the next session will focus on and tailor the document accordingly.

Workflow

  1. Write the handoff document first.

Include the objective, current state, files and commands already touched, blockers, validation state, suggested skills, and next concrete steps. Reference existing PRDs, plans, ADRs, issues, commits, diffs, or artifacts by path or URL instead of duplicating them. Redact secrets and personally identifiable information.

  1. Decide whether the next session needs isolation.

Read [references/worktree-isolation.md](references/worktree-isolation.md) before launching any edit, repair, commit, PR, or parallel-worker handoff. Done when the handoff says whether the worker is read-only or has a dedicated worktree/branch.

  1. Choose fresh session vs fork.

Default to a brand-new Codex session/thread. Read [references/codex-session-choice.md](references/codex-session-choice.md) before creating a Codex app thread or fork. Done when any fork has a written reason that depends on raw inherited conversation history, not convenience.

  1. If running under Codex CLI with tmux, choose placement deliberately.

Read [references/tmux-placement.md](references/tmux-placement.md) before running the bundled tmux helper. If TMUX is empty, do not run tmux commands.

  1. For repair or PR workers, include the repair contract.

Read [references/repair-pr-handoffs.md](references/repair-pr-handoffs.md). Done when the worker is told how to prove the repair, create/update PR proof, run review, and report residual risk.

  1. Launch only after the handoff and placement/isolation decisions are clear.

When creating a brand-new session outside tmux, seed it with only a concise message that links to the handoff document, states the next focus, tells the new agent to read the handoff before acting, and names any required repo/worktree path. For forks, still link the handoff document.

Done Means

  • The handoff document is saved outside the workspace.
  • The document is compact but self-contained enough for a fresh agent to act.
  • Suggested skills are named.
  • Edit/repair/PR/parallel work has a dedicated worktree unless the user

explicitly asked to share the checkout.

  • Read-only shared-checkout work is labeled read-only.
  • Fresh session is used by default; fork is used only for raw-history need.
  • tmux commands run only when TMUX is set.
  • Launch status is reported as started only when a real thread/session/worktree

is verified; queued setup is reported as pending.

Avoid

  • making parallel workers share the coordinator checkout or each other's

worktrees;

  • forking just because the current checkout has uncommitted changes;
  • treating pendingWorktreeId as proof that a worker is running;
  • pointing a repair agent at discovery artifacts as its implementation

workspace;

  • ending a repair handoff with only "fixed it" or a terse file list;
  • pasting secrets, raw environment files, credentials, or unnecessary personal

data into the handoff.

Source & license

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

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

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.