agentleFS
osocietyosociety/vernet/.github/skills/close-chat/SKILL.md

close-chat

Skill566 starsChanged 13 days agoApache-2.0View on GitHub

Close out a chat or work session by preserving durable context in the right memory, updating affected skills and agents when guidance changed, reviewing this skill itself, and giving a concise handoff. Use for close-chat, wrap-up, session closeout, or before ending a Vernet task.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    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
  1. Distinctive:

    Use this skill when the user asks to close, wrap up, or preserve the useful results of the current chat.

    line 9
  2. Distinctive:

    Treat the user's request to close the chat as authorization to make relevant memory and guidance updates, not to commit, push, deploy, …

    line 26
  3. Distinctive:

    Never claim to have ended the chat.

    line 27
  4. Distinctive:

    Do not turn a one-off task outcome into a permanent preference or repository rule without evidence it will matter again.

    line 28
  5. Distinctive:

    Leave unrelated user changes untouched.

    line 29

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “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 Chat

Use this skill when the user asks to close, wrap up, or preserve the useful results of the current chat. It prepares a durable handoff and final response. It does not close the chat session through an API; no such action is available here.

Procedure

  1. Review the current conversation and worktree changes. Identify only verified, durable information that will help in future work: decisions, repository conventions, resolved pitfalls, important unfinished work, and exact validation results. Do not preserve routine chat details or unverified assumptions.
  2. Inspect existing /memories/ files before writing. Update the narrowest applicable memory rather than duplicating guidance:
    • /memories/ for durable user-wide preferences or practices.
    • /memories/repo/ for verified facts and conventions specific to this repository.
    • /memories/session/ only for an unfinished task that must carry into a later chat; do not use it for completed work. Preserve unrelated notes. Never store secrets, credentials, personal data that is not needed, or speculative conclusions.
  3. Decide whether the session revealed a durable change to repository guidance. Update only the relevant skill or agent when its instructions are incomplete, incorrect, or missing a reusable lesson. Keep instructions consistent across overlapping guidance and avoid copying whole sections.
  4. Review this skill itself. If the closeout workflow exposed a reusable gap, update this skill narrowly; do not edit it merely to record the current session or create self-referential update churn.
  5. Verify any edited memory or customization for scope, duplication, and internal consistency. Re-read a memory file immediately before editing it: several agent sessions can share these directories, and a concurrent writer may have already added a related entry, or a contradictory one that needs correcting. For customization files, preserve valid frontmatter and ensure the skill name matches its directory. Do not run application tests for documentation-only closeout edits.
  6. Respond with a brief closeout: what durable information was recorded, which guidance files changed (if any), the task's final state and exact verification status, and any remaining handoff item. Be explicit if there was nothing worth retaining.

Boundaries

  • Treat the user's request to close the chat as authorization to make relevant memory and guidance updates, not to commit, push, deploy, publish, or modify unrelated files.
  • Never claim to have ended the chat; provide the handoff as the final response instead.
  • Do not turn a one-off task outcome into a permanent preference or repository rule without evidence it will matter again.
  • Leave unrelated user changes untouched.

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.