agentleFS
labsiqballabsiqbal/wrap-session/SKILL.md

wrap-session

Skill0 starsChanged 2 months agoMITView on GitHub

Use when wrapping a session with a durable handoff.

3 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions3 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 3

    1. nmamano/isomux★ 43same name
    2. VaibhavAher100/wrap-session★ 1same name
    3. DanielOnPurpose/danielonpurpose-skills★ 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 repository2 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

7 rules. 0 are common, 6 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 6

    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. Distinctive:

    Use before leaving session with code, decisions, documents, operational changes, or unfinished work another session must resume.

    line 18
  2. Very specific:

    Use harness session ID when available.

    line 31
  3. Distinctive:

    Do not write yet. Show:

    line 60
  4. Distinctive:

    Ask owner to approve or edit:

    line 93
  5. Distinctive:

    Run only approved writes and Git actions.

    line 105
  6. Distinctive:

    Create parent directory when missing.

    line 110
  7. Distinctive:

    Run companion detector when available:

    line 153

“Common” is measured: kept by at least 6 in 10 of the 3 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

Wrap Session

Close current session cleanly. Chat is draft. Disk is truth.

/wrap-session means: inspect session residue, propose durable actions, wait for owner approval, execute only approved actions, then write receipt.

Use

Use before leaving session with code, decisions, documents, operational changes, or unfinished work another session must resume.

Skip chitchat with nothing durable.

Receipt

One receipt per session:

~/.local/state/session-wrap/<session_id>.md

Use harness session ID when available. Otherwise generate local-YYYYMMDD-HHMMSS-<cwdslug>.

Receipt front matter must contain:

  • session_id
  • harness: active harness name, or other
  • source: active session surface, or other
  • cwd: absolute path
  • project: nearest Git root name, nearest directory name, or (root)
  • wrapped_at: ISO-8601 timestamp
  • next: one first action

Body headings: Done, Decisions, Written, Git, Deferred, Next.

A receipt means clean close. Unfinished work is deferred, never an open receipt. Re-wrap same session ID overwrites old receipt.

Procedure

1. Inspect

Collect session ID, source, absolute current directory, project label, and—when inside a Git worktree—git status --short --branch.

Completion: every receipt field has a truthful value.

2. Draft Wrap Plan

Do not write yet. Show:

WRAP PLAN
project: ...
cwd: ...

Done:
- outcome + evidence path, commit, or URL

Decisions to persist:
1. decision → proposed path, or skip

Deferred work:
- item → proposed backlog, ticket, or document home

Git:
- clean, or dirty and intentional
- proposed commit message, if needed
- push: no by default

Memory:
- stable fact, or none

Next session starts with:
- one action

Point to existing artifacts. Do not copy whole chat into a document. Put every unfinished item in a real home or explicitly mark it skipped.

3. Owner gate

Ask owner to approve or edit:

  1. Decisions, Deferred, and Memory writes
  2. Git action: leave, commit, or commit+push
  3. Plan edits

Accept all, selected numbers, free-text edits, or cancel.

On cancel: write nothing and report cancellation.

4. Execute approved actions

Run only approved writes and Git actions. Verify each external effect before claiming success. Report failures as failures.

5. Write receipt

Create parent directory when missing. Write:

---
session_id: <id>
harness: <harness>
source: <session surface|other>
cwd: <absolute path>
project: <project>
wrapped_at: <ISO-8601>
next: <one action>
---

# Wrap Session · <project>

## Done
- ...

## Decisions
- ...

## Written
- path — what, or (none)

## Git
- ...

## Deferred
- item → home, or (none)

## Next
- ...

Completion: receipt file exists at expected path.

6. Confirm

Reply with receipt path, files or Git changes made, deferred work and its home, and Next line. Stop.

Optional status report

Run companion detector when available:

bin/wrap-status.sh
bin/wrap-status.sh init

It reads Hermes state.db plus receipts. init is explicit and creates only receipt directory plus rollout baseline. Without Hermes state database, it reports receipts and exits successfully.

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.