agentleFS
aemerson-commitsaemerson-commits/shareable-skills/skills/handoff/SKILL.md

handoff

Skill1 starsChanged 8 months agoMITView on GitHub

Write a disposable session-handoff doc to OS temp so a fresh session/agent can continue ONE in-flight task slice without re-deriving context. Triggers: 'hand this off', 'write a handoff', 'spin up a fresh session to continue X', or when context is getting large and you want a clean continuation. NOT for end-of-day notes (use /session-notes) or cross-session promises (use your memory/notes file).

  • Reads credentials

191 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use decision-mapping, setup-ts-deep-modules and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • FabienDehopreSkill

    decision-mapping

    Turn a loose idea into a sequenced map of investigation tickets, then drive them to resolution one at a time.

    Also in 8 projects that use this

  • liufeng1976Skill

    setup-ts-deep-modules

    Wire dependency-cruiser into a TypeScript repo so each package is a deep module — implementation hidden in subfolders, reachable only through its entry-point files. User-invoked.

    Also in 42 projects that use this

  • AWSSkill

    to-questionnaire

    Turn a decision you can't fully answer into a questionnaire for someone else to fill in.

    Also in 93 projects that use this

  • GoogleSkill

    ask-matt

    Ask which skill or flow fits your situation. A router over the skills in this repo.

    Also in 87 projects that use this

  • RoentekSkill

    loop-me

    Grill me about specs for the workflows I want to build, within this workspace.

    Also in 51 projects that use this

  • LesaniSkill

    writing-beats

    Shape an article as a journey of beats, choose-your-own-adventure style. The user picks a starting beat from the raw material, you write only that beat, then offer options for where to pivot next, beat by beat, until the article reaches a natural end. Use when the user has raw material and wants to assemble it as a narrative rather than an argument.

    Also in 54 projects that use this

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 172 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Redact secrets: Never write a token/key/secret *value* into the handoff.

    line 25
  2. Distinctive:

    Write to the OS temp dir, never the repo or any permanent notes file:

    line 32
  3. Distinctive:

    After writing, report the absolute path and a 2-line summary to the user.

    line 77

“Common” is measured: kept by at least 6 in 10 of the 172 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Handoff

Produce a single disposable handoff document in the OS temp directory that lets a fresh agent or session pick up exactly one in-flight task slice and keep going. Pattern credit: Matt Pocock's "write the handoff, then start clean."

This is not /session-notes (durable, end-of-day) and not your memory/notes file (long-lived deferred items). A handoff is throwaway: it captures live, in-the-weeds state for an immediate continuation and is deleted/ignored once consumed.

When to use

  • Context is growing large and you want to start a clean session that continues the same task.
  • You're deliberately dispatching a fresh agent to carry one slice forward.
  • You're stopping mid-task and a different machine/session will resume soon.

Required: a focus

A handoff covers one task slice, not the whole session. If the user hasn't named the slice, ask: "What single task should this handoff cover?" Do not dump everything you've touched — a handoff that tries to cover three threads is useless to the next agent.

Principles

  1. Pointers, not duplication. The next agent has the repo and git. Give coordinates: path/to/file.js:120-148, commit SHAs, docs/plans/<plan>.md, the relevant session note, the failing test name. Do not paste large code blocks or re-explain what a file already says.
  2. Redact secrets. Never write a token/key/secret value into the handoff. Reference it by name and location (e.g. "API_TOKEN in root .env"). Your scratch/temp dir is less protected than the repo.
  3. Suggested skills. List the skills the continuation will likely need (e.g. /deploy, /test-scheduler, /pre-merge-review) so the fresh agent doesn't rediscover them.
  4. State the exact next action. End with the single concrete next step, not a menu. The next agent should be able to act on line one.
  5. Machine-agnostic. Resolve the temp dir portably (below) — handoffs are written and read on different machines.

Temp path resolution (machine-agnostic)

Write to the OS temp dir, never the repo or any permanent notes file:

# bash / git-bash
TMP="${TMPDIR:-${TEMP:-/tmp}}"
OUT="$TMP/handoff-$(date +%Y%m%d-%H%M%S).md"
# PowerShell
$Out = Join-Path $env:TEMP ("handoff-{0}.md" -f (Get-Date -Format yyyyMMdd-HHmmss))

Document template

# Handoff — <task slice> — <YYYY-MM-DD HH:MM>

## Goal
<one sentence: what "done" looks like for THIS slice>

## State right now
- Branch: <branch> @ <short-sha> (pushed? ahead of origin?)
- <what's working / what's half-built — 2-4 bullets, pointers only>

## Coordinates
- <file path:line> — <why it matters>
- Plan: <docs/plans/...>  |  Session note: <your scratch/temp dir or notes file>
- Commits in play: <sha> <subject>

## Next action
<the single concrete next step>

## Verify
<how to confirm the slice is done — exact command / URL / test>

## Suggested skills
</deploy, /test-scheduler, ...>

## Watch out for
<the one or two gotchas that will bite the next agent — link any relevant feedback/notes files>

Closeout

After writing, report the absolute path and a 2-line summary to the user. Do not commit it, add it to any permanent notes, or write it into memory — it's disposable. If the continuation happens in this same session, just read it back when resuming.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.