end
Skill3 starsChanged 25 days agoMITView on GitHub
Close a work session into a restartable handoff: reconcile relevant repository and Linear state, preserve durable knowledge under the current role's authority, and leave user changes intact. Auto-trigger when the user ends or wraps up a session.
10 other projects use a version of this.
Adopted by
Most recently taken up by
- chris-peterson/tacktheir own version, first seen Oct 26, 2026
- aalishan001/opencockpittheir own version, first seen Oct 25, 2026
- pche-broken-artist/forge-of-thoughttheir own version, first seen Oct 13, 2026
- jantichy/claudetheir own version, first seen Oct 13, 2026
- conorbronsdon/agent-context-ostheir own version, first seen Oct 8, 2026
- princessing/claude-skillstheir own version, first seen Jul 28, 2026
Used with
Projects with it also use update.
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.
AnthropicSkillupdate
Sync tasks and refresh memory from your current activity. Use when pulling new assignments from your project tracker into TASKS.md, triaging stale or overdue tasks, filling memory gaps for unknown people or projects, or running a comprehensive scan to catch todos buried in chat and email.
Also in 3 projects that use this
Other versions10 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 10
- jantichy/claude★ 45same name
- conorbronsdon/agent-context-os★ 30same name
- pche-broken-artist/forge-of-thought★ 12same name
- aalishan001/opencockpit★ 2same name
- bastndev/skills★ 2same name
- conorbronsdon/agent-workspace★ 2same name
- alectivism/maven-template★ 1same name
- princessing/claude-skills★ 1same name
- chris-peterson/tack★ 0same name
- torarnv/claude-project-manager★ 0same name
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 repository16 other files it gives its agents.
- AGENTS.mdAGENTS.md
- assistant-bridge/AGENTS.mdAGENTS.md
- claude/AGENTS.mdAGENTS.md
- codex/AGENTS.mdAGENTS.md
- deepseek/AGENTS.mdAGENTS.md
- kimi/AGENTS.mdAGENTS.md
- big-researchSkill
- cold-emailSkill
- context-saveSkill
- deploySkill
- discourse-atlasSkill
- obscured-siteSkill
- outreach-atlasSkill
- research-vizSkill
- topic-atlasSkill
- triageSkill
The file
We split this skill into the rules it gives an agent
7 rules. 0 are common, 7 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 9 edited versions we read keep these.
7
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 21
Does not count: a generic recap.
- Distinctive: line 23
Stop when the handoff states what is complete and open, every in-scope ledger update has either succeeded or has a named blocker, and git …
- Distinctive: line 27
Read the applicable AGENTS.md, active spec or task brief, relevant Linear issues and comments, the current diff, and git status.
- Distinctive: line 28
Run the smallest verification that supports the outcome.
- Distinctive: line 29
Commit this session's own verified work under the standing authorization.
- Distinctive: line 31
Read the runtime's knowledge-management reference before handling proposals or durable memory.
- Distinctive: line 35
When a check, ledger write, memory judgment, or push is blocked, keep the artifact and report the evidence rather than declaring a clean …
“Common” is measured: kept by at least 6 in 10 of the 9 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
End Session
Goal: leave completed work, remaining work, and their durable records in agreement. Apply only the parts relevant to this session. Ordinary conversation needs no ledger update or handoff; save a brief when unfinished work has context worth carrying forward.
Success means:
- relevant issues have truthful signed outcomes; finished work is moved to Done only after its outcome is recorded;
- a future session can identify the working-tree state, verification evidence, and strongest next action;
- durable knowledge is judged, recorded by its authorized writer, and remains findable;
- pre-existing or another session's changes remain untouched.
Does not count: a generic recap; closing an issue before its outcome is recorded; treating a transient outcome as durable fact; writing frozen daily-note data; or a delegated agent settling OptMem.
Stop when the handoff states what is complete and open, every in-scope ledger update has either succeeded or has a named blocker, and git status has been read again.
Close the loop
- Read the applicable
AGENTS.md, active spec or task brief, relevant Linear issues and comments, the current diff, andgit status. Identify the work this session actually owns; do not claim unrelated dirty files. - Run the smallest verification that supports the outcome. Shared
CLAUDE.mdgoverns independent review; project instructions govern additional checks. Keep evidence with the result, including unrun checks and their reason. - Commit this session's own verified work under the standing authorization. Preserve existing user changes. Push only to a destination and scope already authorized in the active conversation; publication remains ask-first.
- For tracked work, reconcile the live ledger using the runtime's follow-up convention: post a signed outcome on each materially worked issue before moving completed work to Done; file material follow-ups needing ownership or later action; leave failed writes visible and the issue open.
- Read the runtime's knowledge-management reference before handling proposals or durable memory. Apply its current grammar, duplicate checks, proposal dispositions, and verification rules. Record only material outcomes, current state, durable propositions, or find-again documents; never mutate the frozen daily-note layer.
- Authority is role-specific: a user-facing primary judges and writes accepted OptMem records; a delegated agent returns canonical candidates to its primary and never settles OptMem. Honor a field-dispatch result contract when one exists, without inventing one for ordinary work.
- Re-read
git status. Report completed and open work, verification, commits/pushes (or why absent), ledger/memory outcomes, remaining working-tree state, and the exact restart action.
When a check, ledger write, memory judgment, or push is blocked, keep the artifact and report the evidence rather than declaring a clean close.
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.

