handoff
Skill4 starsChanged yesterdayMITView on GitHub
Write a handoff document from the current conversation, so that a new session can continue the work.
- Reads credentials
192 other projects use a version of this.
Adopted by
Most recently taken up by
- tanishq286/tq-forgetheir own version, first seen Nov 15, 2026
- niki914/zafirotheir own version, first seen Nov 15, 2026
- davidondrej/skillstheir own version, first seen Nov 15, 2026
- quickchatai/quickreviewtheir own version, first seen Nov 15, 2026
- HiteSit/real-scientist-skillstheir own version, first seen Nov 15, 2026
- alexj11324/Mikkatheir own version, first seen Nov 15, 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 versions192 projects use a version of this.
Word for word · 1
- andymai/skills★ 4100% same
Edited · 0
None found.
Wrote their own version · 191
- mattpocock/skills★ 325ksame name
- QuivrHQ/quivr★ 40ksame name
- The-Vibe-Company/quivr★ 40ksame name
- GreyDGL/PentestGPT★ 16ksame name
- ykdojo/claude-code-tips★ 10ksame name
- ArcReel/ArcReel★ 6.6ksame name
- vinvcn/mattpocock-skills-zh-CN★ 5.8ksame 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 repository8 other files it gives its agents.
- AGENTS.mdAGENTS.md
- agent-docsSkill
- interviewSkill
- obsidian-projectSkill
- outward-writingSkill
- retrospectiveSkill
- shepherd-prSkill
- ticketsSkill
The file
We split this skill into the rules it gives an agent
6 rules. 0 are common, 5 are distinctive, and 1 is very specific to this project.
0
Common
Most of the 172 edited versions we read keep these.
5
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
- Distinctive: line 10
Write one document that lets an agent with no memory of this conversation continue the work.
- Distinctive: line 15
If that line is empty, write for a session that continues the current task.
- Distinctive: line 22
Repo state: run git status --short and git branch --show-current.
- Distinctive: line 31
Use these sections, in this order.
- Distinctive: line 61
Do not copy content that a file, a commit, a PR, or an issue already holds.
- Very specific: line 73
Read the document once for secrets and personal data: API keys, tokens, passwords, connection strings, email addresses, customer data.
“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
Write one document that lets an agent with no memory of this conversation continue the work. The reader has only the document and the repo.
The focus of the next session, if the user gave one: $ARGUMENTS
If that line is empty, write for a session that continues the current task.
1. Collect the state
Get each item from what you observed in this session, not from memory of what you intended. Run the command when a command can answer.
- Repo state: run
git status --shortandgit branch --show-current. Record the branch, the last commit, and the uncommitted files. - Goal: what the user asked for, in one or two sentences, in the user's words.
- Done: each completed part, with the evidence (the test summary line, the commit).
- In progress: the part that is not complete, and the exact point where it stopped.
- Not started: the parts that remain.
2. Write the document
Use these sections, in this order. Leave out a section that has no content.
# Handoff: <task>
## Goal
## State
## Next step
## Decisions
## Dead ends
## Open questions
## Pointers
## Skills to useRules for each section:
- State: label each claim verified (you ran a command and read the output), inferred (you read code that implies it), or assumed. Quote the summary line of the last test run. A reader acts on "tests pass" as a fact, so do not write it unless you saw it.
- Next step: one action that the reader can start with no more questions. Give the file, the function, and the command. "Continue the refactor" is not a next step.
- Decisions: each decision with its reason, so that the reader does not reopen it.
- Dead ends: each approach that failed, with the error or the reason. This section stops the next session from a repeat of the same attempts.
- Open questions: each question that only the user can answer.
- Pointers: paths, the branch, PR and issue URLs, and the commands to build and test.
- Skills to use: the skills and agents that the next session needs, by full name.
Do not copy content that a file, a commit, a PR, or an issue already holds. Give the path or the URL. A copy goes out of date, and the reader cannot tell which version is correct.
Example of a next step:
- Bad: "Finish the cache work and fix the remaining test."
- Good: "Make
tests/cache/evict.test.tspass. It fails withexpected 3, received 4at line 52, becauseevictOldest()insrc/cache/lru.tsdoes not count the entry it adds. Runpnpm test tests/cache/evict.test.tsto see the failure."
3. Remove secrets
Read the document once for secrets and personal data: API keys, tokens, passwords,
connection strings, email addresses, customer data. Replace each one with a description
of where the reader can get it, such as "the token in .env, key STRIPE_KEY".
4. Save it outside the workspace
Save the document in the temporary folder of the operating system, not in the repo, so
that nobody commits it. Use $TMPDIR if it is set, and /tmp if it is not. On Windows,
use %TEMP%. Name the file handoff-<short-task-name>-<YYYY-MM-DD>-<HHMM>.md, with the
current time. Never overwrite a file: if that name exists, put -2 before .md, then
-3, until the name is free. A second session with the same task must not replace the
notes of the first.
5. Check, then report
- Each path in the document exists. Run
lson each one. - Each "it", "this", and "the bug" names its subject in the same sentence.
- The next step names a file and a command.
- Each claim in State has a label, and each verified claim has its evidence.
- The document has no secret.
Print the full path of the file. Then print one line that the user can paste into a new
session, such as: Read <full path> and continue from "Next step".
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.

