agentleFS
yokeloopyokeloop/yoke/skills/handoff/SKILL.md

handoff

Skill1 starsChanged 4 months agoMITView on GitHub

Saves the live state of the current conversation to `.yoke/handoff/` so a fresh session resumes where this one stopped, referencing existing artifacts instead of duplicating them. Activates when the user writes "handoff", "hand off", "save the context", "dump the conversation", "compact the conversation", "summarize for the next session", "prepare a handoff doc", "pass this to another agent".

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

6 rules. 0 are common, 2 are distinctive, and 4 are 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.

  • 4

    Very specific

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

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

    Write one file to .yoke/handoff/<YYYYMMDD-HHMMSS>-<slug>.md.

    line 25
  2. Very specific:

    Create .yoke/handoff/ when it is missing.

    line 41
  3. Very specific:

    When .yoke/handoff/ already holds a handoff of this same conversation, link it as a prior handoff instead of restating it.

    line 63
  4. Distinctive:

    Never duplicate a PRD, plan, ADR, issue, commit, or diff — link it.

    line 66
  5. Very specific:

    Treat $ARGUMENTS as the focus of the next session and tailor the document to it.

    line 72
  6. Distinctive:

    Delete nothing: handoffs accumulate.

    line 83

“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

Compact this conversation into a document that carries everything the next session needs to continue it. Judge the result by one question: reading only this file and the artifacts it links, could a fresh agent pick the work up without asking what was already agreed?

Capture what lives only in the chat — the open tasks, the user's exact requirements, the alternatives already rejected, the state of the working tree. Everything already written down elsewhere gets a link, never a retelling.

Where to save

Write one file to .yoke/handoff/<YYYYMMDD-HHMMSS>-<slug>.md.

Resolve the path against the root checkout, never the current worktree — /do works inside a worktree and /merge deletes it, taking any uncommitted file with it:

dirname "$(git rev-parse --path-format=absolute --git-common-dir)"

Outside a git repository the root is the working directory; say so in the final message. Take the timestamp from date '+%Y%m%d-%H%M%S' — never invent it.

Derive <slug> from the first source that answers: the active .yoke/ai/<slug>/ task, the current branch name, or a kebab-case summary of the conversation.

Create .yoke/handoff/ when it is missing. When .yoke/ itself was missing, suggest /yoke:bootstrap once the file is written.

What to write

Open with the goal — what this conversation set out to do — then cover:

  • Tasks — every task raised, each marked done, in progress, or pending.
  • Requirements — the user's constraints and prohibitions, quoted verbatim. A paraphrase loses exactly the wording the next agent then violates.
  • State — what is done, what is left, and the next step.
  • Decisions — what was settled and why.
  • Rejected — the alternatives considered and the reason each was dropped, so nobody relitigates them.
  • Git — branch, worktree, uncommitted files, open PR.
  • Gotchas — blockers, traps, and anything that cost time to learn.
  • Artifacts — paths and URLs. Yoke artifacts live under .yoke/: .yoke/ai/<slug>/, .yoke/adr/, .yoke/context.md, .yoke/journal.md.
  • Suggested skills — which yoke skills the next agent should invoke, given where the work sits in the flow (unstarted → /yoke:do; implemented but unreviewed → /yoke:review; PR approved → /yoke:merge).

When .yoke/handoff/ already holds a handoff of this same conversation, link it as a prior handoff instead of restating it.

Never duplicate a PRD, plan, ADR, issue, commit, or diff — link it.

Redact secrets and PII — API keys, passwords, tokens — as [REDACTED]. The file is committed with the rest of .yoke/ unless the project ignores it, so a leaked secret lands in git history.

Treat $ARGUMENTS as the focus of the next session and tailor the document to it.

Finish by printing the absolute path, so the user can hand it to the next agent.

Rules

  • Save under .yoke/handoff/ in the root checkout — never in a worktree, never outside .yoke/.
  • Be exhaustive about what exists only in the conversation; link everything else.
  • Quote the user's requirements verbatim.
  • Redact secrets and PII.
  • Delete nothing: handoffs accumulate.
  • Language: match the conversation language, or follow the project-level definition in CLAUDE.md / AGENTS.md.

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.