agentleFS
cristobaltormocristobaltormo/cortex-md/skills/close-session/SKILL.md

close-session

Skill1 starsChanged 4 days agoMITView on GitHub

Orderly end of a work session. Use it when the user says "closing the session", "that's it for today", "I'm leaving" or similar, and on your own initiative after finishing a big block of work. Leaves memory, pending items, project files and git up to date so the next session does not start blank.

11 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

9 rules. 0 are common, 3 are distinctive, and 6 are very specific to this project.

  • 0

    Common

    Most of the 10 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If a note on the topic exists, fix it.

    line 21
  2. Very specific:

    If it is new: one fact per file, verified set to today, a line in the index of its area.

    line 22
  3. Very specific:

    Link: every new note links with [[ ]] to its project or topic, and the project note links back.

    line 23
  4. Distinctive:

    Do not write what the inventory already knows, or anything readable from the code.

    line 25
  5. Distinctive:

    Add what is waiting for a decision or unfinished.

    line 30
  6. Very specific:

    If you worked on a project, bring its README, changelog, docs and internal notes up to date, without contradictions between them (grep …

    line 35
  7. Very specific:

    Commit and push the brain, and check it went up: git status -sb must not say ahead.

    line 46
  8. Very specific:

    If the memory system itself stumbled for real, one improvement proposal in two or three sentences (rules in memory-keeping).

    line 50
  9. Distinctive:

    If the session was long, suggest continuing in a new one.

    line 52

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

Close the session

The user says it at the end so that nothing stays loose. Do all of it, in order; skip whatever does not apply without commenting.

1. Review what happened

A short list, for yourself: what was done, what was decided and why, what was learned, what is left half done. What only mattered inside the conversation is not saved.

2. Memory

For each decision with its reason, new preference, odd device behaviour or hard-won procedure, apply the memory-keeping skill:

  • If a note on the topic exists, fix it; do not create a second one. Old content goes to ## Before.
  • If it is new: one fact per file, verified set to today, a line in the index of its area.
  • Link: every new note links with [[ ]] to its project or topic, and the project note links back. Do not repeat what another note says; link it.
  • Do not write what the inventory already knows, or anything readable from the code.
  • A repeatable procedure is a new skill, not a note.

3. PENDING.md

Add what is waiting for a decision or unfinished; remove what closed. The file must stay short: detail lives in the note, here only a link and the action.

4. Project files

If you worked on a project, bring its README, changelog, docs and internal notes up to date, without contradictions between them (grep before closing).

5. Cold review

For complex or long work, launch a subagent without the session's context to critically review the documented result. Check what it says and fix what is true. Not needed for small changes.

6. Check and push

  • bin/health.py clean (or say which defect predates this session).
  • Commit and push the brain, and check it went up: git status -sb must not say ahead.

7. Farewell

  • If the memory system itself stumbled for real, one improvement proposal in two or three sentences (rules in memory-keeping).
  • If the session was long, suggest continuing in a new one.
  • Answer in a few lines: what was saved and where, and what is pending. No file dumps.

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.