agentleFS
boettiger-labboettiger-lab/agent-skills/human-communication-consent/SKILL.md

human-communication-consent

Skill4 starsChanged 15 days agoBSD-3-ClauseView on GitHub

Carl's rule that every message to other people goes through his explicit OK first. Read BEFORE opening a pull request or issue, posting a comment or review, requesting reviewers, assigning someone, @-mentioning anyone, sending email or chat messages, or sharing a link with anyone. Draft the text and wait for his OK, or hand it to him to send.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    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:

    Do the technical work: commits, a pushed branch, a built preview.

    line 32
  2. Distinctive:

    Never guess anyone's GitHub handle, email or chat name, and never @-mention someone to get their attention.

    line 40

≈ “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

Messages to other people need Carl's OK first

Anything that other people will see or be notified about goes through Carl first. Either he says OK to the exact text before it is sent, or he sends it himself. Permission to do the technical work is not permission to communicate about it.

What counts

  • Opening a pull request or an issue, including a draft PR, and editing the title or body of one that other people can see.
  • Comments, reviews, review requests, approvals, assignments, labels that notify people, and closing or reopening someone else's PR or issue.
  • Any @-mention, in any text, anywhere.
  • Email, Slack, chat, calendar invitations, and sharing a document or artifact with someone.
  • Automation that does any of these, such as a workflow that comments on PRs. Prefer a check status or job summary over a bot comment, and ask before adding a bot that posts.

These do not need a separate OK when they are part of the task he asked for: local commits, and pushing a feature branch to a remote (not main or another deploy branch).

How to do it

  1. Do the technical work: commits, a pushed branch, a built preview.
  2. Show Carl the exact text you would post (title, body, where it goes) and stop. Ask him whether to send it, or give him the gh command so he can run it himself with ! .
  3. Send it only after he says OK to that text, and only once. An OK covers that message only, not later ones.

Names and handles

  • Never guess anyone's GitHub handle, email or chat name, and never @-mention someone to get their attention. Refer to people by role ("the web developer") unless Carl gives the handle and asks for the mention.

If something was sent by mistake

Tell Carl straight away what was sent, where, and to whom, and whether anyone was notified. Editing or deleting a post does not take back a notification that has already gone out.

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.