handoff
Skill0 starsChanged 18 days agoMITView on GitHub
Use to carry work across time, in four modes. Write: when ending, pausing or switching a session, write a dated handoff note (goals, context, state, checklist, decisions, resume steps) that a fresh session can act on with no chat history. Resume: read the latest note and verify it before continuing. Catch up: in a long chat or unattended run, say in plain words what changed since the user's last message and list every call waiting on them, each with a pick. Recall: when no note exists, rebuild recent working context from past sessions, git and the shared record. Not for a report of yesterday's work for other people (use standup).
191 other projects use a version of this.
Adopted by
Most recently taken up by
- xiaoxiaosrm/dsh-mattpocock-skillstheir own version, first seen Nov 8, 2026
- ethosagent/ethostheir own version, first seen Nov 8, 2026
- dsifry/metaswarmtheir own version, first seen Nov 8, 2026
- glebis/claude-skillstheir own version, first seen Nov 8, 2026
- domengabrovsek/agent-configtheir own version, first seen Nov 8, 2026
- domengabrovsek/claudetheir own version, first seen Nov 8, 2026
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.
FabienDehopreSkilldecision-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
liufeng1976Skillsetup-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
AWSSkillto-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
GoogleSkillask-matt
Ask which skill or flow fits your situation. A router over the skills in this repo.
Also in 87 projects that use this
RoentekSkillloop-me
Grill me about specs for the workflows I want to build, within this workspace.
Also in 51 projects that use this
LesaniSkillwriting-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
Other versions191 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 191
- mattpocock/skills★ 309ksame name
- QuivrHQ/quivr★ 40ksame name
- The-Vibe-Company/quivr★ 40ksame name
- GreyDGL/PentestGPT★ 16ksame name
- ykdojo/claude-code-tips★ 10ksame name
- ArcReel/ArcReel★ 6.3ksame name
- vinvcn/mattpocock-skills-zh-CN★ 5.5ksame name
- withcoral/coral★ 4.9ksame name
- epicenter-md/epicenter★ 4.8ksame name
- EpicenterHQ/epicenter★ 4.8ksame name
181 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository57 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .cursor/rules/comments.mdcCursor rule
- .cursor/rules/commit-messages.mdcCursor rule
- .cursor/rules/git-workflow.mdcCursor rule
- .cursor/rules/manage-stack.mdcCursor rule
- .cursor/rules/no-worktrees.mdcCursor rule
- manage-stackSkill
- convert-as-skillSkill
- create-agentSkill
- create-agents-mdSkill
- create-commandSkill
- create-hookSkill
- create-ruleSkill
- create-skillSkill
- retrospectiveSkill
- super-architectureSkill
- super-cachingSkill
- super-elixirSkill
- super-nodejsSkill
- super-protobufSkill
- super-tech-blueprintSkill
- atomic-semantic-commitSkill
- code-reviewSkill
- create-bruno-collectionSkill
- create-documentationSkill
- create-ticketsSkill
- ready-specSkill
- write-adrSkill
- write-changelogSkill
- write-mr-descriptionSkill
- write-planSkill
- write-postmortemSkill
- write-runbookSkill
- noslopSkill
- output-saversSkill
- safe-refactorSkill
- tddSkill
- verifySkill
- get-movingSkill
- later-listSkill
- standupSkill
- challenge-meSkill
- codebase-learnerSkill
- researchSkill
- second-opinionSkill
- traceSkill
- color-kitSkill
- design-system-kitSkill
- diagramSkill
- infographicSkill
- product-polish-workflowSkill
- prototype-wireframeSkill
- ui-ux-kitSkill
- evidenceSkill
- find-skillsSkill
- rumblingSkill
- use-levistackSkill
The file
We split this skill into the rules it gives an agent
11 rules. 0 are common, 8 are distinctive, and 3 are very specific to this project.
0
Common
Most of the 172 edited versions we read keep these.
8
Distinctive
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 52
Place: a handoffs/ folder at the project root.
- Distinctive: line 55
Run the cold-start check: Read the note as if you had never seen this session:
- Distinctive: line 60
Check for secrets: Write the names of environment variables, never their values.
- Very specific: line 66
If the session exposed repeated friction, offer retrospective after saving the handoff.
- Distinctive: line 70
Use this when the user says "continue", "resume", "pick up", or names a handoff file.
- Distinctive: line 73
Read all of it before you act.
- Distinctive: line 76
Run the test or build command from "Resume" and compare the result.
- Distinctive: line 77
Check that the files and ids it names still exist.
- Distinctive: line 80
When you stop, update the note (status, checklist, state), or write a new dated note that supersedes it.
- Distinctive: line 86
Use absolute dates ("2026-10-04"), not "today" or "yesterday".
- Very specific: line 89
Do not keep stale notes as truth.
“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.
Files
Handoff
Carry work across time. Pick the mode that fits:
| Mode | When | Where |
| Write | You are ending, pausing or switching a session | This file |
| Resume | A dated handoff note exists and the user says "continue" or "pick up" | This file |
| Catch up | The user is back in a long chat or an unattended run: what changed, and what waits on them | catch-up.md |
| Recall | No note exists: rebuild recent context from past sessions, git and the shared record | recall.md |
Catch up and recall are read-only. The two modes below leave a note that lets a fresh session (or a teammate) continue without re-deriving anything. The reader has zero memory of this session. Write for that reader.
A good handoff answers four questions:
- Goals: what are we trying to achieve, and when is it done?
- Context: why does this work exist, and what must the reader know first?
- State: where did we stop, and what is verified?
- Checklist: what is done, and what comes next, in order?
Write a handoff
- Collect facts from the machine, not from memory.
- Date: run
date +%F. Never guess it. - Git:
git branch --show-current,git rev-parse --short HEAD,git status --short. - Tests: use the last result from this session. If you did not run them, write "not run".
- Links: the issue, the pull request or merge request, and any doc the work depends on.
- Date: run
- Pick the file name and place.
- Name:
YYYY-MM-DD-handoff-<topic-slug>.md, for example2026-10-04-handoff-auth-refactor.md. The date prefix makes files sort in time order and shows at a glance how old a note is. - Place: a
handoffs/folder at the project root. Use the project's own notes folder instead when it has one (for exampledocs/handoffs/). Use the exact path when the user gives one, even a fixed name such asHANDOFF.md. - Same topic, same day: update the existing file. Same topic, a later day: write a new file and add
Supersedes: <old path>.
- Name:
- Fill in YYYY-MM-DD-handoff-template.md. See 2026-10-04-handoff-example-export-limit.md for a filled note. Both file names show the convention. Keep it short. Leave out a section only when it is truly empty, and then write "None."
- Run the cold-start check. Read the note as if you had never seen this session:
- Does the first screen say what this is, where work stopped, and what to do next?
- Is every command copy-pasteable, and every path relative to the project root?
- Is there any "as discussed", "see above", or "the bug we found"? Replace each with the actual fact.
- Are decisions recorded with a reason, and are unverified claims marked?
- Check for secrets. Write the names of environment variables, never their values. Never paste tokens, keys, cookies, or customer data.
- Do not commit the note unless the user asks.
Reply with the path, the three-line summary, and a ready prompt for the next session:
Read
<path>. Verify the state it describes. Then continue from the first unchecked item in the checklist.
If the session exposed repeated friction, offer retrospective after saving the handoff.
Resume from a handoff
Use this when the user says "continue", "resume", "pick up", or names a handoff file.
- Find the note. The user's path wins. Otherwise take the newest dated file:
find handoffs -name '*-handoff-*.md' 2>/dev/null | sort | tail -1. The date prefix makes the last name the newest. If it prints nothing, ask the user where the note is. - Read all of it before you act.
- Verify the state. Do not trust it. Compare with reality:
- Branch,
HEADcommit, andgit statusagainst the "State" section. - Run the test or build command from "Resume" and compare the result.
- Check that the files and ids it names still exist.
- Branch,
- Report any drift in two or three lines (for example "HEAD moved by 2 commits", or "the tests fail now"). Ask before you continue if the drift changes the goal.
- Continue from the first unchecked checklist item. Tick items as you finish them.
- When you stop, update the note (status, checklist, state), or write a new dated note that supersedes it.
Pitfalls
- Record decisions and their reasons, not a play-by-play of the session.
- Say what is unverified. "Tests not run" is useful information. A guess written as a fact is harmful.
- Use absolute dates ("2026-10-04"), not "today" or "yesterday".
- Link to long logs, diffs, and docs. Quote only the one decisive line.
- A handoff is not a status report for a manager. Write what the next session needs to act.
- Do not keep stale notes as truth. If you find a note that is wrong, correct it or mark it
Status: superseded.
Related skills
verify: check what is proven before you write the note.retrospective: turn repeated session friction into proposed changes to agent tools.trace: recall uses its source playbooks for the shared record.polyglot-copywriter: catch up and recall are written in plain words.get-moving: resume with one concrete step.later-list: carry open later-list items into the note.standup: a report of yesterday's work for other people. Recall is for your own context, not a report.write-plan: the note says the plan is missing.
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
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.