agentleFS
shasyasanshasyasan/claude-rein/skills/handoff/SKILL.md

handoff

Skill2 starsChanged 11 days agoMITView on GitHub

Read the handoff document to resume work (receiving side), or have the handoff writer update it before requesting a handover (leaving side). Triggers on phrases like 'hand off', 'handoff', 'pick up where I left off', or 'resume'. Also the first step for a session that kickoff pointed at a handoff document.

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

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

  • 0

    Common

    Most of the 172 edited versions we read keep these.

  • 1

    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:

    When 2 and 3 disagree, go with 2 (the effective value).

    line 20
  2. Distinctive:

    Don't send back a long summary of what you read.

    line 35

“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

Receive and leave the handoff document

A rein lineage has exactly one handoff document. The successor session's kickoff points to this one document, and no other. What's written here is only how to decide where it lives and the discipline for reading and writing it -- what actually goes in it (what to write, for which project) is up to whoever writes it.

Receiving side (asked to pick up work, or launched from kickoff)

1. Decide where it lives (check these in order; the first one that resolves wins)

OrderWhat to checkNotes
1The absolute path from your kickoff at launchThe value the watcher resolved at the moment of handover. Most reliable.
2The effective value of rein --cwd <target project> config get handoff_pathReturns the default (next to the records) even for a lineage with nothing configured.
3pointer.handoff_path from rein --cwd <target project> status --jsonThe value the predecessor declared. Can go stale after the document moves, so check this last.

When 2 and 3 disagree, go with 2 (the effective value). rein --cwd <target project> status's handoff: line already shows the effective value, its origin, and whether it exists, all in one place -- checking it once is enough to confirm where you decided it lives.

If it can't be resolved, doesn't exist, or is empty, don't start filling the gap with guesses. Don't go looking for it elsewhere (don't enumerate similarly-named files, working notes, or old logs) -- tell the user where the document was supposed to be and what the situation is, and ask.

2. Read it

  • Start from the three sections "Where things stand," "Next steps," and "Standing decisions" (the standing-decisions section constrains everything that follows -- don't skip it). Open "In flight at handover" when you need to decide whether to reload something. What else to read, and how far, is for the handoff document to decide.
  • This is the only document. Don't go looking for another one on the assumption that one must exist.

3. Reply on your first turn

  • Where things stand (how far the work has gotten)
  • What to do next (the next step the document points to)
  • Whether you're waiting on the user's judgment (and if so, on what)

Don't send back a long summary of what you read. This one reply should be enough to decide whether you need the user's go-ahead before starting work.

4. Check for concurrent work before starting

  • Another session that shares this document may be alive at the same time. Before starting, check who's running with ListAgents, and ask that session directly if your scope looks like it overlaps (the document alone can't tell you what a session running since the handover has already changed).
  • Right after a handover runs, read the handover log (handover.log, sitting with the lineage records). Where the records live isn't fixed -- a --root lineage keeps them outside the target project -- so resolve it the same way as the document itself, from the records: line of rein --cwd <target project> status, rather than assuming .rein/. A bug in the handover mechanism itself is only noticeable from a seat that's actually using it. If something looks off, record it wherever that environment keeps such notes.

Leaving side (before requesting a handover)

  1. Have the handoff writer update the document before rein request. The document is written by rein's handoff writer (the rein:handoff-writer agent), not by you -- how to write it lives in that agent's definition. A handover request is validated on the assumption that the document is already complete, and it's the only thing the successor reads.
  2. Launch the writer with the text rein gives you, exactly as given. rein's handover notices (the usage advisory, the handover trigger, the blocked stop) carry it. If you have no such notice at hand (handing over by hand, before any notice), run rein --cwd <target project> status from this session: its handoff writer: entry prints the same text, with every location filled in. If that entry says it can't, stop and ask the user -- don't build the prompt yourself.
  3. Wait for the writer's report, then place the handover request (see /rein:request). While the writer is running, the request is refused (it's a running child).
  4. If the request is rejected because the document's section structure doesn't match the template, launch the writer again as in step 2, then retry -- don't fix the document yourself.

The document's location follows the same effective value as the receiving side (rein config get handoff_path); if nothing exists yet, rein --cwd <target project> init creates the template.

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.